2,427

samples

2,483

wavelengths

sources

targets

metadata

NIR

family

Provenance & citation

Where the original data and publications come from. Follow these links to obtain the source and cite its authors. Reading guide

Contributor	Robustness of high-throughput prediction of leaf ecophysiological traits using NIRS and poro-fluorometry
Conversion status	partial
Origin · dataverse [open]	10.57745/WVAPOL — Recherche Data Gouv

Governance & integrity

Source license, redistribution conditions, content version and identity hashes. Redistribution tier and original-source availability are separate. Reading guide

Tier	private
License	LicenseRef-not-cleared
Permitted use	Research and benchmarking.
Access policy	Manual download / private-use-only per source.
Redistribution	Dataverse API reports Etalab Open License 2.0.
Content version	1.1.0
Schema / protocol	2.0
Content hash	`fd046afd18be97f7…`
Processing hash	`fdee0227c529160b…`
Metadata hash	`b8297830d48b7931…`

Dataset property explorer

Start with the profile axes to locate unusual variation, then inspect the measured values and evidence below. Scores are descriptive heuristics. Reading guide

Mean available profile score0.44

Highest axisLocal artifacts · 1.00

Diagnostics10

Sources profiled3

How to read the profile

The center is 0 and the outer ring is 1. Larger values indicate stronger anomaly or heterogeneity signals on that axis. Compare individual axes using the table; this is not an overall quality grade. Missing axes are left blank, and partial profiles have no filled polygon.

Profile axes

Dimensionless scores from 0 to 1 summarize each property; larger values indicate stronger signals. Unavailable axes are not zero.

Integrity ? Combines missing cells, zero columns and infinite values. Higher values call for checking data completeness.	0.00
Noise ? Combines roughness and its signal-to-roughness proxy. Real narrow bands may also contribute.	0.01
Local artifacts ? Combines spike, jump and clipping heuristics. Inspect flagged spectral regions before attributing a cause.	1.00
Baseline/shape ? Summarizes baseline slope and excess edge variation relative to signal scale.	0.30
PCA outliers ? Combines relative PCA residual and score-distance tails. Valid rare spectra can produce a high value.	0.49
Distance to average spectrum ? Dispersion around this dataset's mean spectrum, using distance, shape and shift summaries.	0.58
Repeatability ? Dispersion among observations sharing an ID. Interpretation depends on the documented meaning of that ID.	unavailable
Multi-regime structure ? Variation in local density in PCA space. May reflect legitimate batches or subpopulations.	0.71

Possible explanations to investigate

Bars rank heuristic signatures from 0 to 1. They do not estimate the probability of a physical fault. On narrow screens, scroll the chart horizontally. Full labels, scores, evidence coverage, possible causes and required external checks are in the table below.

Heuristic signature scores are not calibrated probabilities of physical causes: 0.78 does not mean a 78% causal probability. Missing evidence remains unavailable. RMS/SAM describe dispersion around the dataset mean, without an independent external reference.

Diagnostic hypotheses and supporting evidence
Diagnostic	Score	Heuristic strength	Evidence coverage	Signals	Possible cause	Required external checks
Signature compatible with a spectral spliceX_asd	0.80	strong	100 %	Spike rate 1.00, Jump rate 1.00, RMS/SAM to mean spectrum 1.00	Break at detector junctions, local calibration or different probe.	Check the white/black reference actually acquired and its metadata. Check instrument, probe, geometry, campaign and processing/resampling log. Check the repeat protocol before interpreting an intra-ID dispersion.
Signature compatible with resamplingX_asd	0.58	medium	100 %	Spike rate 1.00, Jump rate 1.00, High S/N roughness proxy 1.00	Digital artifacts or incorrect spectral processing.	Check the white/black reference actually acquired and its metadata. Check instrument, probe, geometry, campaign and processing/resampling log. Check the repeat protocol before interpreting an intra-ID dispersion.
Signature compatible with a population shiftX_micronir	0.46	medium	100 %	PCA structure 1.00, Mahalanobis / T2 0.61, RMS/SAM to mean spectrum 0.39	Different but physically plausible variety, species, batch or condition.	Check the white/black reference actually acquired and its metadata. Check instrument, probe, geometry, campaign and processing/resampling log. Check the repeat protocol before interpreting an intra-ID dispersion.
Offset compatible with a different white referenceX_asd	0.45	medium	100 %	RMS/SAM to mean spectrum 1.00, local artifacts 1.00, Mahalanobis / T2 0.41	Systematic discrepancy between campaigns, instruments or white reference.	Check the white/black reference actually acquired and its metadata. Check instrument, probe, geometry, campaign and processing/resampling log. Check the repeat protocol before interpreting an intra-ID dispersion.
Signature compatible with a change in geometryX_asd	0.44	medium	100 %	RMS/SAM to mean spectrum 1.00, Mahalanobis / T2 0.41, PCA Q 0.31	Changing illumination, collection, angle or probe-sample distance.	Check the white/black reference actually acquired and its metadata. Check instrument, probe, geometry, campaign and processing/resampling log. Check the repeat protocol before interpreting an intra-ID dispersion.
Multi-regime datasetX_asd	0.42	medium	100 %	RMS/SAM to mean spectrum 1.00, PCA structure 0.55, Mahalanobis / T2 0.41	Mixture of campaigns, operators, batches, setups or spectral subpopulations.	Check the white/black reference actually acquired and its metadata. Check instrument, probe, geometry, campaign and processing/resampling log. Check the repeat protocol before interpreting an intra-ID dispersion.
Signature compatible with a different backgroundX_asd	0.38	weak	100 %	RMS/SAM to mean spectrum 1.00, Mahalanobis / T2 0.41, PCA Q 0.31	Systematic effect of support, white/black, transflectance or measurement environment.	Check the white/black reference actually acquired and its metadata. Check instrument, probe, geometry, campaign and processing/resampling log. Check the repeat protocol before interpreting an intra-ID dispersion.
Signature compatible with partial coverageX_asd	0.33	weak	100 %	RMS/SAM to mean spectrum 1.00, Mahalanobis / T2 0.41, PCA Q 0.31	Partial spot coverage; contribution from the background or support.	Check the white/black reference actually acquired and its metadata. Check instrument, probe, geometry, campaign and processing/resampling log. Check the repeat protocol before interpreting an intra-ID dispersion.

Spectral sources

Each block is a separate instrument or spectral source. Read its signal summary before using PCA to explore variation between observations. Reading guide

Source coverage
Source	Observations	Sample units
ASD dried leaf spectra	2,395	2,394
MicroNIR leaf spectra	2,408	2,408
NeoSpectra leaf spectra	2,116	2,116

2,081 sample units are measured by every source; 17 by a single source. Sources are joined by declared sample identity, never by row position.

ASD dried leaf spectraMicroNIR leaf spectraNeoSpectra leaf spectra

ASD dried leaf spectra

X_asd · NIR · ASD

Read the signal and variation summaries for this source. Axis unit: nm; signal representation: not recorded.

Spectral signal summary

The mean and median describe the typical signal at each channel; shaded bands show the 25th–75th and 5th–95th percentiles across observations. A wider band means greater variation, which can be biological or instrumental.

Sampling

How many observations and spectral channels were exported, and how the wavelength axis is spaced. Reading guide

Wavelengths ? Number of spectral channels. A channel can represent wavelength, wavenumber or another feature axis.	2,101
Axis range ? Lowest and highest recorded spectral-axis coordinates, in the stated unit.	400–2,500 nm
Mean spacing ? Average difference between adjacent spectral-axis coordinates; compare only within the same unit.	1 nm
Grid ? Uniform means adjacent axis spacings are approximately equal. An irregular grid may need care when applying derivatives.	uniform
Observations ? Exported spectral rows in this source, including repeated measurements when present.	2,395

Signal & quality

Signal level and roughness summaries. These depend on the signal representation and are not instrument acceptance tests. Reading guide

Value range ? Minimum and maximum recorded spectral values, in the original signal representation.	-0.0259 – 0.0655
Mean range ? Minimum and maximum of the mean spectrum across spectral channels.	-0.019 – 0.0374
Mean level ? Average signal over observations and channels. The historical key mean_reflectance may also store other signal representations.	0.0002968
Area ? Trapezoidal integral of the mean spectrum over the sorted recorded axis; its unit is signal multiplied by axis unit.	0.6238
PTP ? Peak-to-peak: maximum minus minimum of the mean spectrum. Low values indicate little variation along the mean spectrum.	0.05633
Noise RMS ? Historical roughness sigma proxy from robust second differences. Narrow genuine bands can also raise it; this is not repeat-measured instrument noise.	1.6861e-06
Roughness S/N proxy ? Average absolute signal divided by the roughness proxy. Higher means smoother signal relative to its amplitude, not certified instrument SNR.	1.9e+03
Roughness S/N proxy dB ? 20 log10 of the roughness signal-to-noise proxy. Units are decibels; compare only compatible signal representations.	7e+01 dB
Dynamic range ? Maximum minus minimum of the mean spectrum. Its magnitude depends on the signal representation.	0.0563
Smoothness ? Mean per-spectrum RMS of second differences, in signal units. Smaller values indicate smoother spectra under this protocol.	1.425e-05
Saturated ? Fraction at the observed global maximum; a heuristic indicator, not a comparison with known detector limits.	0.0%
X-outliers ? Number flagged by the qualification outlier method. An unusual spectrum can be valid and should not be discarded automatically.	696

Integrity & artefacts

Missing values, narrow spikes and abrupt jumps to investigate in the original spectra. Reading guide

NaN ratio ? Missing spectral cells divided by all spectral cells, displayed as a percentage.	0.00%
Inf count ? Number of positive or negative infinite spectral values; investigate invalid calculations or export.	0
Zero ratio ? Exactly zero spectral cells divided by all cells. Zero can be a valid measurement and is not automatically missing.	0.00%
Spike count ? Local narrow peaks flagged by the second-difference heuristic; real sharp spectral bands may also be flagged.	210,588
Spike rate ? Fraction of tested second-difference positions flagged as spikes, displayed as a percentage.	4.19%
Jump count ? Adjacent-channel changes flagged by the first-difference heuristic, potentially including detector joins.	232,873
Jump rate ? Fraction of tested adjacent-channel differences flagged as jumps, displayed as a percentage.	4.63%
Clip fraction ? Fraction of finite cells near the observed global minimum or maximum; inspect detector limits before concluding saturation.	0.00%

Shape & reference

Variation relative to this dataset's mean spectrum; the mean is not an independently measured calibration reference. Reading guide

Baseline slope ? Linear slope fitted to the mean spectrum over an axis normalized to 0–1. It describes total baseline change, not change per physical axis unit.	-0.0081454
Curvature RMS ? Median per-spectrum RMS of second differences, summarizing curvature and high-frequency variation.	1.3977e-05
D1 RMS ? Median per-spectrum RMS of adjacent-channel first differences; reflects both real spectral structure and artefacts.	0.00022062
RMS to mean ? Root mean square distance to this dataset's mean spectrum, in signal units.	0.0012547
RMS p95 ? 95th percentile of distances to the mean spectrum: 95% of observations have a distance at or below this value.	0.0024773
SAM to mean ? Spectral Angle Mapper: angle to the mean spectrum, in radians. Smaller angles indicate more similar shapes.	0.1729
SAM p95 ? 95th percentile of spectral angles to the dataset mean, in radians. It is less sensitive to overall amplitude than RMS.	0.34453
Affine offset p95 ? 95th percentile of absolute fitted offsets when matching each spectrum to the dataset mean.	0.0002055
Affine gain p95 Δ ? 95th percentile of the absolute gain deviation from 1 in the fit to the mean spectrum.	0.12361
Affine residual p95 ? 95th percentile of residual RMS after gain and offset correction to the mean spectrum.	0.0023899
Xcorr lag p95 ? 95th percentile of the absolute cross-correlation shift relative to the mean spectrum, in spectral channels.	7

Outliers & repeatability

Unusual spectra and within-ID dispersion. Repeated IDs do not by themselves establish technical repetitions. Reading guide

PCA Q p95/median ? Q (squared prediction error) is variation left outside the retained PCA subspace. The displayed ratio compares its 95th percentile with its median.	2.5
Hotelling T2 p95/median ? Hotelling T² measures distance within the retained PCA subspace, scaled by component variance. Ratio of the 95th percentile to the median.	2.7
Mahalanobis H p95/median ? Distance from the PCA-score center after variance scaling. Ratio of the 95th percentile to the median; large values identify relatively extreme spectra.	1.6
Repeat groups ? Number of IDs with several observations. Source documentation is needed to decide whether these are technical repeats, positions or time points.	1
RMS intra-ID ? Within-ID root mean square deviation from the group's mean spectrum, in signal units.	0
SAM intra-ID ? Within-ID spectral angle to the group's mean spectrum, in radians.	1.4901e-08
CV intra-ID ? Within-ID coefficient of variation: standard deviation relative to absolute mean; unstable near zero signal.	0

Dimensionality (PCA)

Principal component analysis (PCA) compresses correlated wavelengths into components. This describes spectral variation, without using prediction targets. Reading guide

Effective rank ? Entropy-based summary of the variance distribution across components. Larger values indicate variation spread across more dimensions.	4.2
PCs → 95% var ? Number of principal components needed to explain 95% of spectral variance, when reached by the computed projection.	6
PCs → 99% var ? Number of principal components needed to explain 99% of spectral variance, when reached by the computed projection.	14
Top-10 cum. var ? Fraction of spectral variance explained by up to the first ten computed principal components, displayed as a percentage.	98.3%

Computed metric scores 30highest alert 1.00

Value is the measured quantity; alert score is its heuristic mapping to 0–1. Levels are weak below 0.42, medium from 0.42, and strong from 0.72. These are display bands, not calibrated acceptance thresholds. A dash means unavailable, never zero. Possible causes are suggestions to investigate. Use each metric’s ? control for its meaning and scoring: hover for a hint, or expand it with a tap, Enter or Space.

Family	Computed metric	Value	Alert score	Heuristic level	Descriptive interpretation	Possible causes to verify	Calculation / scoring
Data integrity	Invalid/masked cell count (owner report) ? Count of cells reported unavailable by the native owner under its null or validity mask. This count is distinct from the historical NaN ratio; consult the recorded population and denominator. Scoring: No alert mapping is defined for this metric; an unavailable score does not mean a zero measured value. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.missing_cells	—	—	not calculable	Not enough information to score this metric on this dataset.	Missing data or validity mask; not evidence of an acquisition cause	Native Preflight count of cells unavailable under x/null and valid_mask; denominator stays in native_metric.denominators.n_cells. Not legacy NaN ratio.
Data integrity	NaN ratio ? Missing spectral cells divided by all spectral cells, displayed as a percentage. Scoring: alert = min(1, nan_ratio / 0.05). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.nan_ratio	0%	0.00	weak	Full spectrum	Acquisition/export error	count(isnan(X)) / X.sizealert = min(1, nan_ratio / 0.05)
Data integrity	Inf count ? Number of positive or negative infinite spectral values; investigate invalid calculations or export. Scoring: alert = min(1, inf_count / 1). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.inf_count	0	0.00	weak	Normal	Invalid calculations	count(isinf(X))alert = min(1, inf_count / 1)
Data integrity	Zero ratio ? Exactly zero spectral cells divided by all cells. Zero can be a valid measurement and is not automatically missing. Scoring: alert = min(1, zero_ratio / 0.05). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.zero_ratio	0%	0.00	weak	Normal	Export, saturation	count(X == 0) / count(finite X)alert = min(1, zero_ratio / 0.05)
Overall amplitude	Mean reflectance ? Average signal over observations and channels. The historical key mean_reflectance may also store other signal representations. Scoring: alert reuses baseline/shape drift because absolute reflectance ranges are technology-dependent. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.mean_reflectance	0.00029685	0.29	weak	Too dark	Background, geometry	mean(X finite)alert reuses baseline/shape drift because absolute reflectance ranges are technology-dependent
Overall amplitude	Area under curve ? Trapezoidal integral of the mean spectrum over the sorted recorded axis; its unit is signal multiplied by axis unit. Scoring: alert reuses baseline/shape drift because area scale depends on axis and units. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.area_under_curve	0.62377	0.29	weak	Normal	Probe distance	trapezoid(mean_spectrum, spectral_axis)alert reuses baseline/shape drift because area scale depends on axis and units
Overall amplitude	Peak-to-peak (PTP) ? Peak-to-peak: maximum minus minimum of the mean spectrum. Low values indicate little variation along the mean spectrum. Scoring: alert increases when dynamic range is abnormally flat. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.peak_to_peak	0.056333	0.00	weak	Strong variability	Saturation	max(mean_spectrum) - min(mean_spectrum)alert increases when dynamic range is abnormally flat
Overall amplitude	Variance ? Variance of all finite spectral cells, in squared signal units. It combines variation across observations and channels; a small value can describe a flat signal. Scoring: alert increases when variance/dynamic range is abnormally flat. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.variance	4.9294e-05	0.00	weak	Normal or heterogeneous	Bad contact	var(X finite)alert increases when variance/dynamic range is abnormally flat
Noise	Roughness sigma proxy ? Historical roughness sigma proxy from robust second differences. Narrow genuine bands can also raise it; this is not repeat-measured instrument noise. Scoring: alert = noise_rms / signal_scale, saturated at 5%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. noise.roughness_sigma_proxy	1.6861e-06	0.00	weak	Stable	Lamp, detector	median MAD(second derivative) * 1.4826 / sqrt(6)alert = noise_rms / signal_scale, saturated at 5%
Noise	Roughness S/N proxy ? Average absolute signal divided by the roughness proxy. Higher means smoother signal relative to its amplitude, not certified instrument SNR. Scoring: alert decreases with SNR dB; >=40 dB is treated as low alert. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. noise.roughness_signal_to_noise_proxy	1927.7	0.00	weak	Good signal	Acquisition	mean(abs(X)) / noise_rmsalert decreases with SNR dB; >=40 dB is treated as low alert
Noise	Bandwise roughness S/N proxy ? Minimum channel-wise ratio of absolute mean signal to local second-difference roughness. A smaller ratio identifies a weak or rough spectral region; genuine narrow bands and near-zero signal can also lower it. This is not certified instrument SNR. Scoring: alert decreases with worst-band SNR dB; >=35 dB is treated as low alert. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. noise.bandwise_roughness_signal_to_noise_proxy_min	0.43281	1.00	strong	Problem area	Detector	min(abs(mean_spectrum) / local second-derivative noise)alert decreases with worst-band SNR dB; >=35 dB is treated as low alert
Local artifacts	Spike count ? Local narrow peaks flagged by the second-difference heuristic; real sharp spectral bands may also be flagged. Scoring: alert follows spike_rate, saturated at 1%. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.spike_count	210,588	1.00	strong	Artifacts	Cosmic rays, splice	count robust outliers in second derivativealert follows spike_rate, saturated at 1%
Local artifacts	Spike rate ? Fraction of tested second-difference positions flagged as spikes, displayed as a percentage. Scoring: alert = min(1, spike_rate / 0.01). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.spike_rate	4.19%	1.00	strong	Suspicious spectrum	Interpolation	spike_count / (n_samples * (n_features - 2))alert = min(1, spike_rate / 0.01)
Local artifacts	Jump count ? Adjacent-channel changes flagged by the first-difference heuristic, potentially including detector joins. Scoring: alert follows jump_rate, saturated at 1%. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.jump_count	232,873	1.00	strong	Detector connection	Splice	count robust outliers in first derivativealert follows jump_rate, saturated at 1%
Local artifacts	Jump rate ? Fraction of tested adjacent-channel differences flagged as jumps, displayed as a percentage. Scoring: alert = min(1, jump_rate / 0.01). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.jump_rate	4.63%	1.00	strong	Spectral problem	Calibration	jump_count / (n_samples * (n_features - 1))alert = min(1, jump_rate / 0.01)
Local artifacts	Clip fraction ? Fraction of finite cells near the observed global minimum or maximum; inspect detector limits before concluding saturation. Scoring: alert = min(1, clip_fraction / 0.01). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.clip_fraction	3.97e-05%	0.00	weak	Normal	Detector saturated	fraction of finite cells equal to repeated min/max extremaalert = min(1, clip_fraction / 0.01)
Spectral form	Baseline slope ? Linear slope fitted to the mean spectrum over an axis normalized to 0–1. It describes total baseline change, not change per physical axis unit. Scoring: alert = abs(slope / signal_scale), saturated at 0.5. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. shape.baseline_slope	-0.0081454	0.29	weak	Stable	Illumination	linear slope of mean_spectrum over normalized axisalert = abs(slope / signal_scale), saturated at 0.5
Spectral form	Curvature RMS ? Median per-spectrum RMS of second differences, summarizing curvature and high-frequency variation. Scoring: alert = curvature_rms / signal_scale, saturated at 1%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. shape.curvature_rms	1.3977e-05	0.02	weak	Smooth	Background, splice	median RMS(second derivative per spectrum)alert = curvature_rms / signal_scale, saturated at 1%
Spectral form	D1 RMS ? Median per-spectrum RMS of adjacent-channel first differences; reflects both real spectral structure and artefacts. Scoring: alert = d1_rms / signal_scale, saturated at 5%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. shape.d1_rms	0.00022062	0.08	weak	Flat	Biology or artifact	median RMS(first derivative per spectrum)alert = d1_rms / signal_scale, saturated at 5%
Multivariate outliers	PCA Q (SPE) ? Q (squared prediction error) is variation left outside the retained PCA subspace. The displayed ratio compares its 95th percentile with its median. Scoring: alert = min(1, pca_q_ratio / 8). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. outliers.pca_q_ratio	2.4643	0.31	weak	Compliant	Artifact, mixture	p95(Q/SPE residual) / median(Q/SPE residual)alert = min(1, pca_q_ratio / 8)
Multivariate outliers	Hotelling T² ? Hotelling T² measures distance within the retained PCA subspace, scaled by component variance. Ratio of the 95th percentile to the median. Scoring: alert = min(1, hotelling_t2_ratio / 8). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. outliers.hotelling_t2_ratio	2.6689	0.33	weak	Central	Natural variability	p95(Hotelling T2) / median(Hotelling T2)alert = min(1, hotelling_t2_ratio / 8)
Multivariate outliers	Mahalanobis H ? Distance from the PCA-score center after variance scaling. Ratio of the 95th percentile to the median; large values identify relatively extreme spectra. Scoring: alert = min(1, mahalanobis_h_ratio / 4). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. outliers.mahalanobis_h_ratio	1.6337	0.41	weak	Normal population	Different domain	p95(sqrt(T2)) / median(sqrt(T2))alert = min(1, mahalanobis_h_ratio / 4)
Comparison to average spectrum	RMS to mean spectrum ? 95th percentile of distances to the mean spectrum: 95% of observations have a distance at or below this value. Scoring: alert = RMS_p95 / signal_scale, saturated at 25%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. reference.rms_to_mean_spectrum_p95	0.0024773	0.18	weak	Typical	Domain shift	p95 RMS distance to dataset mean spectrumalert = RMS_p95 / signal_scale, saturated at 25%
Comparison to average spectrum	Spectral Angle Mapper (SAM) ? 95th percentile of spectral angles to the dataset mean, in radians. It is less sensitive to overall amplitude than RMS. Scoring: alert = min(1, SAM_p95 / 0.35 rad). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. reference.sam_to_mean_spectrum_p95	0.34453	0.98	strong	Different shape	Background, geometry	p95 spectral angle to dataset mean spectrumalert = min(1, SAM_p95 / 0.35 rad)
Repeatability	RMS intra-ID ? Within-ID root mean square deviation from the group's mean spectrum, in signal units. Repeated IDs require documentation before this spread can be called technical repeatability. Scoring: alert = RMS_intra_ID / signal_scale, saturated at 10%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. repeatability.rms_intra_id	0	0.00	weak	Stable	Positioning	median RMS distance to repeated-sample centroidalert = RMS_intra_ID / signal_scale, saturated at 10%
Repeatability	SAM intra-ID ? Within-ID spectral angle to the group's mean spectrum, in radians. Repeated IDs require documentation before this spread can be called technical repeatability. Scoring: alert = min(1, SAM_intra_ID / 0.15 rad). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. repeatability.sam_intra_id	1.4901e-08	0.00	weak	Stable	Acquisition	median SAM to repeated-sample centroidalert = min(1, SAM_intra_ID / 0.15 rad)
Repeatability	CV intra-ID ? Within-ID coefficient of variation: standard deviation relative to absolute mean; unstable near zero signal. Repeated IDs require documentation before this spread can be called technical repeatability. Scoring: alert = min(1, CV_intra_ID / 0.25). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. repeatability.cv_intra_id	0	0.00	weak	Stable	Operator	median within-ID band CValert = min(1, CV_intra_ID / 0.25)
Structure of the dataset	PCA score density ? Inverse median distance to nearby observations in retained PCA-score space. Larger raw values mean more closely spaced observations; scale depends on the PCA representation. The alert uses density variation, so raw density alone does not establish an anomaly. Scoring: alert follows density_cv/profile structure complexity, not raw density alone. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. structure.pca_score_density	44.611	0.55	medium	Subpopulations	Different batches	1 / median kNN distance in PCA score spacealert follows density_cv/profile structure complexity, not raw density alone
Structure of the dataset	kNN distance ratio (not LOF) ? 95th percentile of mean k-nearest-neighbor distance divided by the population median of that distance, in PCA-score space. Values above 1 indicate an upper tail with more isolated observations; valid rare spectra can also be isolated. This distance ratio is not the Local Outlier Factor algorithm. Scoring: alert = min(1, max(0, kNN_distance_ratio_p95 - 1) / 2). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. structure.knn_distance_ratio_p95	1.6518	0.33	weak	Normal population	Rare cases	p95(mean kNN distance / median population mean kNN distance) in PCA-score space; not LOFalert = min(1, max(0, kNN_distance_ratio_p95 - 1) / 2)
Structure of the dataset	Isolation Forest score ? 95th percentile of the negated Isolation Forest score in retained PCA-score space. Larger raw scores indicate easier isolation under that model; the scale is implementation-dependent. The displayed alert reuses structure complexity rather than directly mapping this raw score. Scoring: alert follows structure complexity; raw score is implementation-dependent. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. structure.isolation_forest_score_p95	0.5368	0.55	medium	Atypical spectrum	Various causes	p95 IsolationForest anomaly score on PCA scoresalert follows structure complexity; raw score is implementation-dependent

PCA score plot — explore spectral variation

Source ASD dried leaf spectra: 800 of 2,395 source observations are plotted. PC1 captures 56.6% and PC2 19.9% of the reported spectral variance: 76.5% together; 23.5% remains outside this view.

Principal component analysis (PCA) summarizes spectral differences without using targets. The horizontal PC1 axis captures the largest variation; vertical PC2 captures the next largest direction of variation. Each dot is one observation; repeated measurements need not be distinct samples. Scores are projection coordinates, not wavelengths or target values. Nearby dots have similar projections; separate groups or isolated dots are leads to inspect in the original spectra and acquisition metadata. Overlapping dots can differ in omitted dimensions. The dots have no class or batch labels, so this view cannot identify groups, confirm defective observations, explain a chemical cause or measure prediction accuracy.

View plotted PCA coordinates

Point numbers identify rows in this displayed coordinate list, not original observation or sample IDs. Non-integer coordinates use up to five significant digits. Original IDs, component loadings, preprocessing details and the observation selection method are not supplied with this plot.

Plotted observation positions (not original sample IDs)
Point	PC1 score	PC2 score
1	-0.095702	-0.067926
2	-0.009072	-0.099675
3	-0.076435	-0.0053583
4	-0.046435	0.0021625
5	-0.046435	0.0021625
6	0.046837	0.010769
7	0.046837	0.010769
8	-0.023144	0.002998
9	0.029091	-0.006125
10	-0.04172	-0.039388
11	0.0088892	-0.02889
12	0.014588	0.01714
13	-0.079975	-0.02579
14	-0.091129	-0.014086
15	-0.02606	0.022075
16	0.036356	0.017367
17	0.032294	0.0033819
18	0.011309	0.020923
19	-0.015384	0.020119
20	-0.024497	-0.019513
21	-0.042434	0.029367
22	0.012191	0.027534
23	0.02215	0.026894
24	0.0062789	0.024685
25	-0.012164	-0.024861
26	0.010947	-0.0062654
27	0.043004	0.028899
28	-0.027721	-0.02734
29	-0.0090601	0.012605
30	-0.046881	0.025067
31	-0.076641	-0.013085
32	0.013982	-0.005307
33	0.043394	-0.049602
34	0.03851	-0.02288
35	-0.030861	-0.036138
36	0.013166	-0.0014053
37	0.0017171	0.0029609
38	0.023518	0.022908
39	0.015273	0.00188
40	-0.0022507	-0.016708
41	0.033341	-0.0066991
42	0.051001	0.030974
43	0.021729	-0.053536
44	-0.033604	-0.051692
45	0.010979	0.006761
46	0.016799	0.020018
47	0.072497	0.012042
48	0.058461	0.0092042
49	0.047478	0.015819
50	0.055659	-0.001208
51	0.048556	0.012876
52	0.011214	-0.027583
53	0.036313	0.026861
54	-0.014156	-0.010242
55	-0.035291	-0.021378
56	0.00042475	-0.032327
57	0.039417	0.00075302
58	0.017196	-0.026314
59	0.038596	0.006121
60	0.063626	-0.011601
61	-0.031537	0.011342
62	0.062851	-0.025156
63	-0.0823	-0.019622
64	0.017976	-0.0040648
65	-0.0027868	-0.0017711
66	0.029101	0.0050405
67	-0.002789	0.022406
68	0.021305	0.039806
69	0.042168	0.0039775
70	0.028822	-0.0037451
71	0.046603	0.0022002
72	0.039127	-0.0057582
73	0.061539	-0.025132
74	0.040256	0.020119
75	0.049744	0.010276
76	0.075224	-0.0041467
77	0.032781	0.017498
78	0.064106	-0.019755
79	0.04475	-0.035949
80	0.078113	0.00020856
81	-0.017469	0.0091981
82	0.077456	0.020415
83	0.083619	-0.028291
84	0.059233	-0.030119
85	0.053126	0.028257
86	0.026948	0.026434
87	0.071378	0.005886
88	0.051416	0.011753
89	0.058108	-0.014671
90	-0.058794	-0.0011091
91	-0.026874	-0.058198
92	0.069186	-0.012754
93	0.01026	0.011914
94	0.068043	-0.036925
95	-0.02239	0.010809
96	0.049212	-0.017719
97	0.07363	0.01285
98	0.06416	0.001898
99	0.032687	-0.036155
100	0.080974	0.015112
101	-0.11477	-0.047059
102	0.021401	-0.014182
103	0.063026	-0.037798
104	0.05266	-0.013608
105	0.041282	-0.0049306
106	-0.016262	-0.038626
107	0.038217	-0.053804
108	-0.065488	-0.012031
109	0.064456	-0.031903
110	0.068583	-0.010505
111	-0.093724	0.011479
112	-0.027071	0.0012063
113	0.080811	-0.026716
114	0.0088363	-0.040992
115	0.031755	-0.0074351
116	0.064563	-0.023555
117	0.056565	-0.02176
118	0.023262	0.00032583
119	0.026159	0.0081459
120	0.088307	-0.055363
121	0.033951	-0.022674
122	-0.0029626	-0.0060172
123	0.0045976	0.019818
124	-0.017181	0.018626
125	0.035111	0.0055929
126	0.028694	0.0037113
127	0.067895	-0.046221
128	0.061828	-0.016517
129	0.036598	0.0027704
130	0.098926	-0.0034974
131	-0.011416	-0.018286
132	0.057155	-0.0019841
133	0.033192	-0.0018946
134	0.013755	0.046727
135	0.016791	-0.025148
136	0.028886	-0.029862
137	0.052936	-0.015394
138	-0.044418	0.018333
139	0.053949	0.02161
140	0.05704	0.032912
141	0.043851	0.040689
142	0.019287	0.021799
143	0.10605	-0.03573
144	0.076914	-0.017872
145	0.0086421	-0.061987
146	0.027382	-0.037432
147	-0.0044057	-0.02642
148	0.064904	-0.010794
149	0.037599	0.014628
150	0.091462	-0.047164
151	0.052708	-0.018287
152	0.078044	-0.015234
153	0.032222	-0.0099223
154	0.038095	0.023865
155	0.028579	-0.0088063
156	0.059412	-0.03243
157	-0.0025358	-0.077224
158	0.054928	-0.010207
159	-0.0035578	-0.048487
160	0.040871	-0.020146
161	0.027497	-0.064479
162	-0.059609	0.037151
163	0.0038882	0.01968
164	0.060238	0.022217
165	-0.0021064	0.022984
166	0.028532	0.0095684
167	0.032481	0.024894
168	-0.025432	0.022717
169	0.045031	0.019499
170	0.043905	0.0071388
171	0.0079438	0.011583
172	0.060167	-0.0099923
173	0.010659	-0.010222
174	0.035506	-0.0058118
175	0.0080939	-0.04911
176	0.031945	0.038035
177	-0.014368	0.055154
178	0.10047	-0.013644
179	-0.015787	-3.8227e-05
180	-0.051517	-0.021263
181	0.050058	-0.00059108
182	0.00031798	0.03269
183	-0.0050227	-0.02847
184	0.063818	0.0038491
185	0.078787	0.03019
186	0.070111	0.026402
187	0.017044	-0.016894
188	0.0059332	-0.042943
189	0.0098175	-0.030872
190	-0.021562	0.023982
191	0.044217	0.020209
192	0.01033	0.033469
193	0.046429	0.006164
194	-0.00038929	-0.0093322
195	0.038076	-0.0072341
196	-0.0026799	0.013027
197	-0.00035314	-0.043828
198	0.02603	-0.071022
199	-0.013878	-0.017872
200	-0.14129	-0.034844
201	-0.01706	0.014356
202	-0.0090811	0.027394
203	0.054475	-0.014309
204	0.056644	-0.020996
205	0.080626	-0.027618
206	0.0041474	0.012291
207	0.00068363	0.022198
208	0.052569	0.016358
209	0.063777	0.011944
210	0.08307	0.03016
211	-0.017758	-0.0011602
212	0.10348	0.012113
213	0.06142	-0.00076756
214	0.074676	-0.036921
215	0.035846	-0.027048
216	-0.027214	0.02229
217	0.013688	0.037855
218	-0.0082531	0.013276
219	0.053666	0.030906
220	-0.0085178	0.032706
221	0.027544	0.033854
222	0.026114	0.0044829
223	0.038083	-0.0020994
224	0.065191	-0.051655
225	0.026662	0.0079145
226	0.041003	0.017926
227	0.035267	0.024114
228	0.064854	0.01914
229	0.072041	-0.028933
230	0.10621	-0.054907
231	0.090142	0.025819
232	0.070605	0.031327
233	0.047035	0.0086923
234	0.017662	0.012427
235	0.061403	0.025575
236	0.015485	-0.011939
237	0.029881	0.0048258
238	0.030552	0.015523
239	0.0086674	0.020795
240	-0.0083179	-0.041668
241	0.085258	0.019574
242	0.092961	-0.0084446
243	0.048094	0.0067998
244	-0.0076253	0.028604
245	0.012985	-0.012205
246	0.029099	-0.0012418
247	0.050663	-0.0052531
248	0.057902	0.0078871
249	0.096576	0.013633
250	0.11498	-0.042553
251	0.085415	-0.018234
252	0.051682	0.0026182
253	0.062019	0.04002
254	0.061436	0.032429
255	0.051343	0.0042987
256	0.016101	-0.0095785
257	0.073926	-0.0088635
258	0.0040772	-0.0046724
259	-0.014671	0.0070086
260	-0.092845	-0.025835
261	0.03457	-0.0087988
262	-0.032367	0.0055668
263	-0.012086	-0.065848
264	-0.054224	-0.0085021
265	0.064234	-0.0082737
266	0.076704	-0.012922
267	0.036153	0.03499
268	0.042201	-0.0017818
269	0.061273	0.0051816
270	0.034877	0.0038464
271	0.032072	0.0049424
272	-0.011838	0.018361
273	0.048542	-0.0011843
274	-0.044071	-0.0038053
275	0.068136	0.036706
276	0.010812	-0.001604
277	0.0088174	-0.023211
278	-0.079485	0.0142
279	-0.05517	0.037334
280	0.018718	0.00081469
281	0.025326	0.0021965
282	0.024851	-0.0031947
283	-0.009441	-0.022878
284	0.022145	-0.034956
285	0.031661	0.011722
286	0.02169	-0.012545
287	0.00040181	0.0087277
288	0.017796	-0.035173
289	0.00235	-0.0095754
290	0.0349	-0.029714
291	-0.0084984	-0.004042
292	0.065081	-0.014663
293	0.056431	-0.012864
294	0.038479	-0.023991
295	0.053282	-0.00031725
296	-0.013157	0.012859
297	0.088033	-0.0054851
298	-0.011319	-0.014597
299	0.050465	-0.0046421
300	0.015985	-0.00055102
301	0.016375	0.0082964
302	-0.037542	-0.011169
303	0.0061514	0.035998
304	0.089529	0.013182
305	0.015036	-0.0033628
306	0.071341	-0.027627
307	0.11294	-0.010641
308	0.10153	-0.027506
309	0.031475	-0.0021942
310	0.021715	0.0081746
311	0.02765	0.016753
312	0.048862	-0.0035734
313	0.048529	-0.03316
314	0.10904	-0.013461
315	0.070779	-0.006223
316	0.1203	-0.018318
317	0.05286	0.0096959
318	-0.0071852	0.03139
319	0.014962	0.008382
320	0.058182	-0.017863
321	0.005628	-0.0097637
322	0.020649	0.050682
323	0.11697	0.010711
324	-0.046875	0.014807
325	-0.020945	0.0026156
326	0.069095	-0.015783
327	0.047905	0.015737
328	0.038594	0.01911
329	0.11519	-0.065685
330	0.1067	-0.070637
331	0.12103	-0.050101
332	-0.012194	0.036711
333	0.050248	0.034729
334	-0.079584	0.00015439
335	-0.032454	0.016669
336	0.021759	-0.076395
337	-0.063375	-0.0091118
338	-0.0097031	0.046328
339	-0.054648	0.030654
340	0.031332	0.0090067
341	3.7433e-05	0.022905
342	-0.044952	-0.033213
343	0.0336	0.01781
344	-0.10508	0.015801
345	-0.013767	0.034136
346	-0.00093201	0.0072819
347	0.054847	-0.010711
348	0.0099582	-0.016494
349	0.012819	0.020942
350	0.036489	0.020057
351	-0.016238	0.046172
352	-0.045677	0.023475
353	-0.059153	0.0064961
354	-0.038125	-0.0056142
355	-0.015472	-0.014641
356	-0.077805	0.032122
357	-0.0078097	0.015869
358	-0.046637	0.015814
359	-0.1098	-0.0070168
360	0.075547	0.034413
361	0.015347	0.036437
362	-0.04838	0.0018151
363	-0.063883	0.03308
364	0.077554	-0.0066583
365	0.049428	0.010463
366	0.01133	0.021911
367	0.00098896	0.012473
368	0.040952	0.007767
369	0.03194	0.025249
370	0.01321	-0.0079393
371	0.083318	0.031183
372	0.0070554	-0.0018098
373	0.0091471	-0.022344
374	0.051729	0.02738
375	0.016009	0.0074223
376	-0.031688	-0.010428
377	-0.035773	0.035405
378	0.048637	-0.019733
379	0.054718	-0.0093894
380	0.010546	0.0092341
381	-0.068674	-0.0045027
382	0.10304	-0.040352
383	0.10946	-0.038778
384	0.037771	-0.024226
385	0.0043659	-0.0053139
386	-0.0047937	-0.024981
387	0.017433	-0.024593
388	-0.01086	-0.029921
389	0.0031212	0.029496
390	0.032034	-0.019258
391	-0.031014	-0.00064565
392	0.034523	-0.076116
393	0.086026	-0.02799
394	0.038456	-0.020959
395	0.060635	0.017117
396	0.046381	-0.0079585
397	-0.017592	0.0089113
398	-0.012184	-0.0036201
399	0.039031	-0.010665
400	-0.054022	-0.11967
401	-0.0030808	-0.028467
402	0.041501	-0.028716
403	0.056903	0.03443
404	0.0077847	0.032895
405	0.040265	0.033059
406	0.0080056	0.046085
407	-0.024586	-0.01102
408	-0.02134	-0.013711
409	-0.063086	-0.0024581
410	-0.0043004	0.022942
411	0.024209	0.017465
412	-0.042302	-0.013343
413	0.039788	0.029492
414	-0.033042	0.0087652
415	0.038259	0.015947
416	0.02543	0.015723
417	0.044585	0.037964
418	0.051341	0.035372
419	-0.10519	0.020895
420	-0.074556	0.035217
421	-0.024985	-0.0048526
422	-0.067692	0.038042
423	-0.084913	0.026627
424	-0.069943	0.031068
425	-0.068414	0.012514
426	0.00058339	-0.0041901
427	-0.10476	-0.042648
428	-0.049532	0.024843
429	0.027995	0.022216
430	-0.084322	-0.017053
431	-0.012509	0.021985
432	-0.027838	0.065349
433	-0.04712	0.002499
434	-0.058071	0.056858
435	0.0090534	0.030477
436	-0.012517	-0.038706
437	-0.0057885	0.03918
438	-0.047583	0.0033913
439	0.063277	0.024607
440	-0.029124	0.058708
441	0.0321	0.024213
442	-0.070525	0.016155
443	-0.066569	-0.0064518
444	0.094195	0.018266
445	-0.087327	-0.0011685
446	-0.074114	0.02433
447	-0.03577	0.0047173
448	-0.0015009	0.016504
449	0.017237	0.022483
450	-0.015351	0.0081227
451	-0.0011391	-0.0097654
452	0.023059	0.054968
453	0.031459	0.002197
454	-0.096183	0.0063704
455	0.048394	0.027162
456	-0.041303	3.3307e-05
457	0.020138	-0.020359
458	0.012163	-0.0094531
459	-0.026805	0.035033
460	-0.037992	0.02345
461	0.030207	0.008469
462	-0.034793	0.018991
463	-0.044607	-0.1925
464	-0.0075086	-0.019581
465	-0.13709	-0.055786
466	-0.0019754	0.01276
467	-0.10063	-0.0086821
468	0.032927	0.053656
469	0.03324	-0.0032568
470	0.014914	-0.019088
471	0.0064332	-0.035136
472	-0.022031	-0.059627
473	2.2982e-06	0.018977
474	-0.15585	0.0035753
475	-0.067863	0.038899
476	-0.091089	0.028913
477	-0.12986	-0.12493
478	-0.078974	-0.040532
479	-0.13605	-0.088598
480	-0.11025	-0.040135
481	-0.044865	0.01101
482	0.021987	0.042572
483	0.02405	0.024617
484	-0.034251	-0.01948
485	0.0085512	0.0065065
486	-0.078706	0.0083264
487	-0.052339	0.0093309
488	-0.064317	-0.026834
489	-0.040033	-0.0061165
490	-0.06805	0.028563
491	-0.014692	0.026612
492	-0.052479	-0.011863
493	-0.026727	-0.032164
494	-0.13162	-0.079081
495	-0.018821	-0.0048189
496	-0.061337	0.0051064
497	-0.11316	-0.04524
498	-0.028816	0.0018874
499	-0.10374	-0.026367
500	-0.060578	0.02074
501	-0.059279	0.065386
502	-0.016555	0.015643
503	-0.076386	-0.028095
504	-0.14049	-0.034702
505	0.025437	0.0076286
506	-0.035513	0.0053252
507	-0.017023	0.043292
508	-0.010759	0.066097
509	-0.01134	0.059373
510	0.02637	0.035112
511	-0.048797	0.036074
512	-0.035779	0.034291
513	-0.061351	-0.021709
514	0.0080569	0.026889
515	-0.010732	0.054196
516	-0.0078762	0.015855
517	-0.0036877	0.022482
518	-0.053419	0.021807
519	-0.061909	-0.002952
520	-0.080693	-0.095411
521	-0.016491	-0.041674
522	0.017683	-0.045724
523	0.015931	-0.060527
524	-0.089809	-0.016909
525	0.014811	-0.011777
526	-0.10262	0.0080403
527	-0.010442	0.025127
528	-0.059613	-0.010751
529	-0.10956	-0.045393
530	0.040036	0.019811
531	0.024547	-0.060838
532	-0.0069679	-0.036866
533	-0.087797	0.0018534
534	-0.042798	0.035637
535	0.018127	0.066218
536	-0.10037	0.03666
537	-0.11932	0.025742
538	-0.10528	0.025902
539	-0.055486	0.033976
540	0.077162	0.0050685
541	-0.041134	0.011891
542	-0.042001	0.015941
543	-0.0138	0.013725
544	0.00033865	0.03427
545	0.043016	0.022893
546	-0.076522	0.015576
547	0.0055071	0.023271
548	0.065218	0.036464
549	0.051705	0.018461
550	0.020753	0.013003
551	-0.088922	0.015998
552	-0.009702	0.031344
553	-0.019063	0.027621
554	-0.041587	0.069268
555	0.0016229	0.048511
556	0.044867	-0.0028864
557	0.0071021	0.011101
558	-0.095006	-0.0089291
559	-0.031994	0.023595
560	-0.086181	-0.071598
561	-0.10264	0.0094383
562	-0.050854	0.030289
563	-0.048516	-0.02797
564	-0.042423	0.014005
565	0.01792	-0.090102
566	-0.047939	0.05327
567	-0.013503	0.034637
568	-0.036257	-0.077597
569	-0.077773	0.021333
570	-0.032765	-0.11768
571	-0.037195	0.034258
572	-0.059984	0.030143
573	-0.023405	-0.061771
574	-0.046322	0.0045278
575	-0.039585	0.057672
576	-0.043501	-0.025559
577	0.01278	0.0028968
578	-0.034093	-0.057525
579	0.0026813	-0.016161
580	-0.077675	0.050052
581	-0.014798	0.017069
582	-0.066545	-0.0049446
583	-0.031875	-0.0044014
584	-0.091873	-0.047445
585	-0.060393	0.011797
586	-0.017376	-0.041603
587	-0.051647	-0.039904
588	0.056007	0.037985
589	-0.0081676	0.039464
590	0.037035	0.0027221
591	0.069267	0.0063421
592	-0.068632	0.013124
593	-0.064803	0.0023673
594	-0.04513	0.073961
595	-0.061205	-0.0034045
596	-0.091369	-0.030774
597	-0.12351	0.017472
598	-0.0082039	0.026368
599	0.010455	-0.00059122
600	0.0015901	0.021173
601	-0.10163	-0.045553
602	0.0089665	0.023313
603	0.058819	0.022083
604	-0.041824	0.024082
605	-0.0053323	-0.00043151
606	0.091149	0.011424
607	-0.034149	-0.019425
608	0.039578	-0.049031
609	-0.016686	-0.0021553
610	0.012127	-0.016478
611	-0.049435	-0.035833
612	0.0067206	-0.014036
613	-0.077216	-0.053031
614	0.066061	-0.0075962
615	-0.069157	-0.020899
616	-0.04085	0.020757
617	-0.024009	0.014469
618	0.034885	-0.034234
619	0.033592	-0.024542
620	-0.016413	0.0088349
621	0.019999	-0.057499
622	-0.01179	-0.017781
623	-0.088628	-0.052287
624	-0.09521	-0.062632
625	-0.047468	-0.079466
626	-0.064302	-0.060123
627	-0.026435	-0.043007
628	-0.019545	-0.0039588
629	-0.057521	-0.015107
630	-0.020583	-0.0097482
631	0.037622	-0.0049183
632	-0.034488	-0.036261
633	-0.062452	0.016862
634	-0.0069477	0.004409
635	-0.074983	-0.038154
636	-0.027633	-0.011611
637	0.032916	0.004517
638	-0.056826	0.04166
639	-0.064002	0.0086289
640	-0.088584	0.0146
641	-0.0027056	-0.0085711
642	-0.0070629	-0.018341
643	-0.053831	0.025965
644	0.0056621	-0.001259
645	0.005406	-0.0048036
646	-0.0872	-0.039032
647	0.024262	-0.0004733
648	0.025234	0.010267
649	0.0083264	0.019565
650	0.035888	-0.037413
651	-0.0028922	0.0028803
652	-0.048086	-0.045974
653	-0.061578	-0.036539
654	-0.017072	-0.092585
655	0.05116	-0.01337
656	-0.011822	-0.023178
657	-0.012571	-0.009018
658	-0.060982	0.016916
659	0.0051955	0.020924
660	-0.012603	0.010968
661	-0.020345	0.017215
662	-0.11326	0.051091
663	-0.0026852	0.024703
664	-0.08725	-0.015963
665	-0.078643	0.021924
666	-0.072857	0.048778
667	0.065874	0.0070955
668	0.018198	0.023069
669	-0.10137	0.069309
670	-0.043304	0.0059375
671	-0.0088512	0.021675
672	0.030307	0.00089923
673	0.053842	0.01746
674	-0.049589	0.058504
675	0.068284	-0.0016956
676	-0.026251	0.040306
677	-0.023003	0.022159
678	0.11665	-0.010702
679	0.11827	-0.011345
680	-0.091632	-0.015719
681	0.026369	0.018283
682	-0.031561	0.035754
683	0.048867	0.021674
684	-0.070759	-0.0020926
685	-0.010427	0.030551
686	-0.069102	0.050037
687	-0.034696	0.063784
688	-0.0065748	0.035866
689	-0.05272	0.053059
690	-0.029963	0.031168
691	0.1019	-0.0038649
692	-0.024874	0.038938
693	-0.074068	-0.011529
694	-0.05283	0.025517
695	-0.029208	0.016058
696	0.055336	0.010048
697	-0.086923	0.070453
698	0.01867	0.011956
699	-0.062827	0.0038558
700	0.026467	0.028978
701	-0.072445	0.043045
702	0.035078	-0.08586
703	-0.029224	-0.042133
704	-0.030248	-0.0053557
705	-0.022446	-0.036247
706	-0.094834	-0.0092084
707	0.033997	0.0069876
708	-0.04915	-0.025757
709	-0.030459	-0.022426
710	0.019466	-0.01846
711	-0.030919	-0.01535
712	-0.014648	-0.019039
713	-0.03106	0.025592
714	-0.015625	0.010504
715	-0.012979	-0.0048409
716	0.0051095	-0.068225
717	0.012716	-0.030164
718	-0.08518	0.029576
719	-0.031694	-0.027839
720	-0.048565	0.00094546
721	-0.02886	-0.030897
722	-0.018143	-0.0015812
723	-0.0033275	-0.025704
724	0.035127	-0.017768
725	0.03276	-0.0019173
726	-0.012138	0.022905
727	-0.016865	-0.057686
728	-0.012823	-0.054146
729	-0.019098	0.0033451
730	-0.025467	0.015611
731	-0.0038158	0.028781
732	-0.0028263	0.037007
733	-0.022611	0.045579
734	-0.00078743	0.019759
735	0.052934	0.048442
736	-0.0086899	0.029189
737	0.035344	-0.012799
738	0.015705	0.0013631
739	-0.010896	0.036118
740	-0.11198	-0.022326
741	-0.048602	-0.020496
742	-0.064545	-0.0046448
743	-0.01035	0.011521
744	-0.0097738	0.03792
745	0.019224	0.018196
746	-0.0034211	0.045201
747	-0.0093313	0.032838
748	-0.084717	-0.068497
749	-0.10542	0.04722
750	-0.013218	0.051424
751	-0.057878	-0.0084408
752	-0.035676	-0.0096555
753	0.00057503	0.0084562
754	0.015049	0.019045
755	0.088223	-0.0012463
756	0.0389	-0.0028168
757	-0.060438	-0.013968
758	0.027025	0.0060647
759	0.083367	-0.035397
760	0.017064	0.0037372
761	-0.0049639	-0.014914
762	-0.024225	-0.055301
763	0.009141	-0.05733
764	-0.010207	-0.0070014
765	0.036011	-0.051201
766	0.04292	0.0051573
767	0.018466	-0.0043587
768	0.053896	0.013831
769	0.0022054	0.012298
770	0.048941	0.00042782
771	0.01214	0.012832
772	-0.016491	-0.039479
773	0.023394	-0.0060318
774	0.015958	0.0042408
775	0.027835	0.021273
776	0.007793	0.022412
777	0.01743	-0.013175
778	-0.011337	0.027466
779	0.0052106	0.025031
780	0.001726	0.0022781
781	-0.077184	0.059475
782	-0.019869	0.045105
783	-0.079133	0.01015
784	-0.042824	0.019009
785	0.045881	0.036463
786	0.018537	0.042443
787	0.032861	0.042837
788	-0.013726	-0.026699
789	-0.055639	-0.01081
790	-0.088125	-0.0096647
791	-0.07418	0.017712
792	-0.033256	-0.0029976
793	0.059514	0.044344
794	-0.035179	0.059602
795	-0.022705	-0.0041657
796	-0.051889	0.055297
797	-0.058957	0.01033
798	-0.060654	0.031161
799	-0.053511	-0.010777
800	-0.13747	0.0058516

PCA explained variance — how many dimensions?

Each bar reports the share of spectral variance explained by a component. Concentration in the first components indicates correlated wavelengths. This is not variance explained in the prediction target. This projection may use a different observation sample from the score plot.

View explained variance values

Variance per displayed principal component
Component	Spectral variance
PC1	55.62%
PC2	20.23%
PC3	8.77%
PC4	6.12%
PC5	3.18%
PC6	2.02%
PC7	0.83%
PC8	0.67%
PC9	0.49%
PC10	0.32%

X-Y spectral correlation 7

Pearson r measures the linear association between each spectral channel (X) and a numeric target (Y), from −1 to +1. The table reports peak and average absolute correlation and the share of channels with |r| ≥ 0.5. These are in-dataset associations, not validation scores or proof of causation.

Spectral association with LMA

Read Pearson correlation (−1 to +1) along the spectral axis (nm); the solid curve shows |r| and the dashed curve shows signed r.

Spectral association with WCf

Read Pearson correlation (−1 to +1) along the spectral axis (nm); the solid curve shows |r| and the dashed curve shows signed r.

Spectral association with WQ

Read Pearson correlation (−1 to +1) along the spectral axis (nm); the solid curve shows |r| and the dashed curve shows signed r.

Target	max \|r\|	axis @ max	mean \|r\|	\|r\| ≥ .5
LMA	0.894	1,342	0.318	22.7%
WCf	0.639	1,799	0.284	16.5%
WQ	0.608	1,341	0.221	8.6%
An	0.448	506	0.142	0.0%
PSI	0.336	1,439	0.123	0.0%
WUEintr	0.412	2,450	0.0993	0.0%
WUEinst	0.515	1,197	0.169	0.2%

MicroNIR leaf spectra

X_micronir · NIR · MicroNIR

Read the signal and variation summaries for this source. Axis unit: nm; signal representation: not recorded.

Spectral signal summary

Sampling

How many observations and spectral channels were exported, and how the wavelength axis is spaced. Reading guide

Wavelengths ? Number of spectral channels. A channel can represent wavelength, wavenumber or another feature axis.	125
Axis range ? Lowest and highest recorded spectral-axis coordinates, in the stated unit.	908.1–1676 nm
Mean spacing ? Average difference between adjacent spectral-axis coordinates; compare only within the same unit.	6.19 nm
Grid ? Uniform means adjacent axis spacings are approximately equal. An irregular grid may need care when applying derivatives.	uniform
Observations ? Exported spectral rows in this source, including repeated measurements when present.	2,408

Signal & quality

Signal level and roughness summaries. These depend on the signal representation and are not instrument acceptance tests. Reading guide

Value range ? Minimum and maximum recorded spectral values, in the original signal representation.	-0.0424 – 0.05
Mean range ? Minimum and maximum of the mean spectrum across spectral channels.	-0.0331 – 0.0288
Mean level ? Average signal over observations and channels. The historical key mean_reflectance may also store other signal representations.	-0.0004734
Area ? Trapezoidal integral of the mean spectrum over the sorted recorded axis; its unit is signal multiplied by axis unit.	-0.4563
PTP ? Peak-to-peak: maximum minus minimum of the mean spectrum. Low values indicate little variation along the mean spectrum.	0.06191
Noise RMS ? Historical roughness sigma proxy from robust second differences. Narrow genuine bands can also raise it; this is not repeat-measured instrument noise.	4.9573e-05
Roughness S/N proxy ? Average absolute signal divided by the roughness proxy. Higher means smoother signal relative to its amplitude, not certified instrument SNR.	1.5e+02
Roughness S/N proxy dB ? 20 log10 of the roughness signal-to-noise proxy. Units are decibels; compare only compatible signal representations.	4e+01 dB
Dynamic range ? Maximum minus minimum of the mean spectrum. Its magnitude depends on the signal representation.	0.0619
Smoothness ? Mean per-spectrum RMS of second differences, in signal units. Smaller values indicate smoother spectra under this protocol.	0.0004319
Saturated ? Fraction at the observed global maximum; a heuristic indicator, not a comparison with known detector limits.	0.0%
X-outliers ? Number flagged by the qualification outlier method. An unusual spectrum can be valid and should not be discarded automatically.	1,096

Integrity & artefacts

Missing values, narrow spikes and abrupt jumps to investigate in the original spectra. Reading guide

NaN ratio ? Missing spectral cells divided by all spectral cells, displayed as a percentage.	0.00%
Inf count ? Number of positive or negative infinite spectral values; investigate invalid calculations or export.	0
Zero ratio ? Exactly zero spectral cells divided by all cells. Zero can be a valid measurement and is not automatically missing.	0.00%
Spike count ? Local narrow peaks flagged by the second-difference heuristic; real sharp spectral bands may also be flagged.	14,961
Spike rate ? Fraction of tested second-difference positions flagged as spikes, displayed as a percentage.	5.05%
Jump count ? Adjacent-channel changes flagged by the first-difference heuristic, potentially including detector joins.	2,174
Jump rate ? Fraction of tested adjacent-channel differences flagged as jumps, displayed as a percentage.	0.73%
Clip fraction ? Fraction of finite cells near the observed global minimum or maximum; inspect detector limits before concluding saturation.	0.00%

Shape & reference

Variation relative to this dataset's mean spectrum; the mean is not an independently measured calibration reference. Reading guide

Baseline slope ? Linear slope fitted to the mean spectrum over an axis normalized to 0–1. It describes total baseline change, not change per physical axis unit.	-0.0046698
Curvature RMS ? Median per-spectrum RMS of second differences, summarizing curvature and high-frequency variation.	0.0004302
D1 RMS ? Median per-spectrum RMS of adjacent-channel first differences; reflects both real spectral structure and artefacts.	0.0015956
RMS to mean ? Root mean square distance to this dataset's mean spectrum, in signal units.	0.0006792
RMS p95 ? 95th percentile of distances to the mean spectrum: 95% of observations have a distance at or below this value.	0.0015759
SAM to mean ? Spectral Angle Mapper: angle to the mean spectrum, in radians. Smaller angles indicate more similar shapes.	0.059098
SAM p95 ? 95th percentile of spectral angles to the dataset mean, in radians. It is less sensitive to overall amplitude than RMS.	0.13747
Affine offset p95 ? 95th percentile of absolute fitted offsets when matching each spectrum to the dataset mean.	0.00029575
Affine gain p95 Δ ? 95th percentile of the absolute gain deviation from 1 in the fit to the mean spectrum.	0.061121
Affine residual p95 ? 95th percentile of residual RMS after gain and offset correction to the mean spectrum.	0.0014934
Xcorr lag p95 ? 95th percentile of the absolute cross-correlation shift relative to the mean spectrum, in spectral channels.	0

Outliers & repeatability

Unusual spectra and within-ID dispersion. Repeated IDs do not by themselves establish technical repetitions. Reading guide

PCA Q p95/median ? Q (squared prediction error) is variation left outside the retained PCA subspace. The displayed ratio compares its 95th percentile with its median.	5
Hotelling T2 p95/median ? Hotelling T² measures distance within the retained PCA subspace, scaled by component variance. Ratio of the 95th percentile to the median.	4.9
Mahalanobis H p95/median ? Distance from the PCA-score center after variance scaling. Ratio of the 95th percentile to the median; large values identify relatively extreme spectra.	2.2
Repeat groups ? Number of IDs with several observations. Source documentation is needed to decide whether these are technical repeats, positions or time points.	0

Dimensionality (PCA)

Principal component analysis (PCA) compresses correlated wavelengths into components. This describes spectral variation, without using prediction targets. Reading guide

Effective rank ? Entropy-based summary of the variance distribution across components. Larger values indicate variation spread across more dimensions.	2.2
PCs → 95% var ? Number of principal components needed to explain 95% of spectral variance, when reached by the computed projection.	2
PCs → 99% var ? Number of principal components needed to explain 99% of spectral variance, when reached by the computed projection.	5
Top-10 cum. var ? Fraction of spectral variance explained by up to the first ten computed principal components, displayed as a percentage.	99.9%

Computed metric scores 30highest alert 1.00

Family	Computed metric	Value	Alert score	Heuristic level	Descriptive interpretation	Possible causes to verify	Calculation / scoring
Data integrity	Invalid/masked cell count (owner report) ? Count of cells reported unavailable by the native owner under its null or validity mask. This count is distinct from the historical NaN ratio; consult the recorded population and denominator. Scoring: No alert mapping is defined for this metric; an unavailable score does not mean a zero measured value. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.missing_cells	—	—	not calculable	Not enough information to score this metric on this dataset.	Missing data or validity mask; not evidence of an acquisition cause	Native Preflight count of cells unavailable under x/null and valid_mask; denominator stays in native_metric.denominators.n_cells. Not legacy NaN ratio.
Data integrity	NaN ratio ? Missing spectral cells divided by all spectral cells, displayed as a percentage. Scoring: alert = min(1, nan_ratio / 0.05). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.nan_ratio	0%	0.00	weak	Full spectrum	Acquisition/export error	count(isnan(X)) / X.sizealert = min(1, nan_ratio / 0.05)
Data integrity	Inf count ? Number of positive or negative infinite spectral values; investigate invalid calculations or export. Scoring: alert = min(1, inf_count / 1). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.inf_count	0	0.00	weak	Normal	Invalid calculations	count(isinf(X))alert = min(1, inf_count / 1)
Data integrity	Zero ratio ? Exactly zero spectral cells divided by all cells. Zero can be a valid measurement and is not automatically missing. Scoring: alert = min(1, zero_ratio / 0.05). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.zero_ratio	0%	0.00	weak	Normal	Export, saturation	count(X == 0) / count(finite X)alert = min(1, zero_ratio / 0.05)
Overall amplitude	Mean reflectance ? Average signal over observations and channels. The historical key mean_reflectance may also store other signal representations. Scoring: alert reuses baseline/shape drift because absolute reflectance ranges are technology-dependent. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.mean_reflectance	-0.00047338	0.27	weak	Too dark	Background, geometry	mean(X finite)alert reuses baseline/shape drift because absolute reflectance ranges are technology-dependent
Overall amplitude	Area under curve ? Trapezoidal integral of the mean spectrum over the sorted recorded axis; its unit is signal multiplied by axis unit. Scoring: alert reuses baseline/shape drift because area scale depends on axis and units. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.area_under_curve	-0.45631	0.27	weak	Normal	Probe distance	trapezoid(mean_spectrum, spectral_axis)alert reuses baseline/shape drift because area scale depends on axis and units
Overall amplitude	Peak-to-peak (PTP) ? Peak-to-peak: maximum minus minimum of the mean spectrum. Low values indicate little variation along the mean spectrum. Scoring: alert increases when dynamic range is abnormally flat. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.peak_to_peak	0.06191	0.00	weak	Strong variability	Saturation	max(mean_spectrum) - min(mean_spectrum)alert increases when dynamic range is abnormally flat
Overall amplitude	Variance ? Variance of all finite spectral cells, in squared signal units. It combines variation across observations and channels; a small value can describe a flat signal. Scoring: alert increases when variance/dynamic range is abnormally flat. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.variance	0.00011406	0.00	weak	Normal or heterogeneous	Bad contact	var(X finite)alert increases when variance/dynamic range is abnormally flat
Noise	Roughness sigma proxy ? Historical roughness sigma proxy from robust second differences. Narrow genuine bands can also raise it; this is not repeat-measured instrument noise. Scoring: alert = noise_rms / signal_scale, saturated at 5%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. noise.roughness_sigma_proxy	4.9573e-05	0.02	weak	Stable	Lamp, detector	median MAD(second derivative) * 1.4826 / sqrt(6)alert = noise_rms / signal_scale, saturated at 5%
Noise	Roughness S/N proxy ? Average absolute signal divided by the roughness proxy. Higher means smoother signal relative to its amplitude, not certified instrument SNR. Scoring: alert decreases with SNR dB; >=40 dB is treated as low alert. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. noise.roughness_signal_to_noise_proxy	146.54	0.00	weak	Good signal	Acquisition	mean(abs(X)) / noise_rmsalert decreases with SNR dB; >=40 dB is treated as low alert
Noise	Bandwise roughness S/N proxy ? Minimum channel-wise ratio of absolute mean signal to local second-difference roughness. A smaller ratio identifies a weak or rough spectral region; genuine narrow bands and near-zero signal can also lower it. This is not certified instrument SNR. Scoring: alert decreases with worst-band SNR dB; >=35 dB is treated as low alert. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. noise.bandwise_roughness_signal_to_noise_proxy_min	3.7725	0.67	medium	Problem area	Detector	min(abs(mean_spectrum) / local second-derivative noise)alert decreases with worst-band SNR dB; >=35 dB is treated as low alert
Local artifacts	Spike count ? Local narrow peaks flagged by the second-difference heuristic; real sharp spectral bands may also be flagged. Scoring: alert follows spike_rate, saturated at 1%. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.spike_count	14,961	1.00	strong	Artifacts	Cosmic rays, splice	count robust outliers in second derivativealert follows spike_rate, saturated at 1%
Local artifacts	Spike rate ? Fraction of tested second-difference positions flagged as spikes, displayed as a percentage. Scoring: alert = min(1, spike_rate / 0.01). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.spike_rate	5.05%	1.00	strong	Suspicious spectrum	Interpolation	spike_count / (n_samples * (n_features - 2))alert = min(1, spike_rate / 0.01)
Local artifacts	Jump count ? Adjacent-channel changes flagged by the first-difference heuristic, potentially including detector joins. Scoring: alert follows jump_rate, saturated at 1%. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.jump_count	2,174	0.73	strong	Detector connection	Splice	count robust outliers in first derivativealert follows jump_rate, saturated at 1%
Local artifacts	Jump rate ? Fraction of tested adjacent-channel differences flagged as jumps, displayed as a percentage. Scoring: alert = min(1, jump_rate / 0.01). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.jump_rate	0.728%	0.73	strong	Spectral problem	Calibration	jump_count / (n_samples * (n_features - 1))alert = min(1, jump_rate / 0.01)
Local artifacts	Clip fraction ? Fraction of finite cells near the observed global minimum or maximum; inspect detector limits before concluding saturation. Scoring: alert = min(1, clip_fraction / 0.01). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.clip_fraction	0.000664%	0.00	weak	Normal	Detector saturated	fraction of finite cells equal to repeated min/max extremaalert = min(1, clip_fraction / 0.01)
Spectral form	Baseline slope ? Linear slope fitted to the mean spectrum over an axis normalized to 0–1. It describes total baseline change, not change per physical axis unit. Scoring: alert = abs(slope / signal_scale), saturated at 0.5. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. shape.baseline_slope	-0.0046698	0.15	weak	Stable	Illumination	linear slope of mean_spectrum over normalized axisalert = abs(slope / signal_scale), saturated at 0.5
Spectral form	Curvature RMS ? Median per-spectrum RMS of second differences, summarizing curvature and high-frequency variation. Scoring: alert = curvature_rms / signal_scale, saturated at 1%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. shape.curvature_rms	0.0004302	0.69	medium	Unusual shape	Background, splice	median RMS(second derivative per spectrum)alert = curvature_rms / signal_scale, saturated at 1%
Spectral form	D1 RMS ? Median per-spectrum RMS of adjacent-channel first differences; reflects both real spectral structure and artefacts. Scoring: alert = d1_rms / signal_scale, saturated at 5%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. shape.d1_rms	0.0015956	0.52	medium	Structured spectrum	Biology or artifact	median RMS(first derivative per spectrum)alert = d1_rms / signal_scale, saturated at 5%
Multivariate outliers	PCA Q (SPE) ? Q (squared prediction error) is variation left outside the retained PCA subspace. The displayed ratio compares its 95th percentile with its median. Scoring: alert = min(1, pca_q_ratio / 8). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. outliers.pca_q_ratio	5.0369	0.63	medium	Atypical spectrum	Artifact, mixture	p95(Q/SPE residual) / median(Q/SPE residual)alert = min(1, pca_q_ratio / 8)
Multivariate outliers	Hotelling T² ? Hotelling T² measures distance within the retained PCA subspace, scaled by component variance. Ratio of the 95th percentile to the median. Scoring: alert = min(1, hotelling_t2_ratio / 8). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. outliers.hotelling_t2_ratio	4.9196	0.61	medium	Extreme but consistent	Natural variability	p95(Hotelling T2) / median(Hotelling T2)alert = min(1, hotelling_t2_ratio / 8)
Multivariate outliers	Mahalanobis H ? Distance from the PCA-score center after variance scaling. Ratio of the 95th percentile to the median; large values identify relatively extreme spectra. Scoring: alert = min(1, mahalanobis_h_ratio / 4). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. outliers.mahalanobis_h_ratio	2.218	0.55	medium	Global outlier	Different domain	p95(sqrt(T2)) / median(sqrt(T2))alert = min(1, mahalanobis_h_ratio / 4)
Comparison to average spectrum	RMS to mean spectrum ? 95th percentile of distances to the mean spectrum: 95% of observations have a distance at or below this value. Scoring: alert = RMS_p95 / signal_scale, saturated at 25%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. reference.rms_to_mean_spectrum_p95	0.0015759	0.10	weak	Typical	Domain shift	p95 RMS distance to dataset mean spectrumalert = RMS_p95 / signal_scale, saturated at 25%
Comparison to average spectrum	Spectral Angle Mapper (SAM) ? 95th percentile of spectral angles to the dataset mean, in radians. It is less sensitive to overall amplitude than RMS. Scoring: alert = min(1, SAM_p95 / 0.35 rad). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. reference.sam_to_mean_spectrum_p95	0.13747	0.39	weak	Similar	Background, geometry	p95 spectral angle to dataset mean spectrumalert = min(1, SAM_p95 / 0.35 rad)
Repeatability	RMS intra-ID ? Within-ID root mean square deviation from the group's mean spectrum, in signal units. Repeated IDs require documentation before this spread can be called technical repeatability. Scoring: alert = RMS_intra_ID / signal_scale, saturated at 10%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. repeatability.rms_intra_id	—	—	not calculable	Not enough information to score this metric on this dataset.	Positioning	median RMS distance to repeated-sample centroidalert = RMS_intra_ID / signal_scale, saturated at 10%
Repeatability	SAM intra-ID ? Within-ID spectral angle to the group's mean spectrum, in radians. Repeated IDs require documentation before this spread can be called technical repeatability. Scoring: alert = min(1, SAM_intra_ID / 0.15 rad). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. repeatability.sam_intra_id	—	—	not calculable	Not enough information to score this metric on this dataset.	Acquisition	median SAM to repeated-sample centroidalert = min(1, SAM_intra_ID / 0.15 rad)
Repeatability	CV intra-ID ? Within-ID coefficient of variation: standard deviation relative to absolute mean; unstable near zero signal. Repeated IDs require documentation before this spread can be called technical repeatability. Scoring: alert = min(1, CV_intra_ID / 0.25). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. repeatability.cv_intra_id	—	—	not calculable	Not enough information to score this metric on this dataset.	Operator	median within-ID band CValert = min(1, CV_intra_ID / 0.25)
Structure of the dataset	PCA score density ? Inverse median distance to nearby observations in retained PCA-score space. Larger raw values mean more closely spaced observations; scale depends on the PCA representation. The alert uses density variation, so raw density alone does not establish an anomaly. Scoring: alert follows density_cv/profile structure complexity, not raw density alone. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. structure.pca_score_density	1313.9	1.00	strong	Subpopulations	Different batches	1 / median kNN distance in PCA score spacealert follows density_cv/profile structure complexity, not raw density alone
Structure of the dataset	kNN distance ratio (not LOF) ? 95th percentile of mean k-nearest-neighbor distance divided by the population median of that distance, in PCA-score space. Values above 1 indicate an upper tail with more isolated observations; valid rare spectra can also be isolated. This distance ratio is not the Local Outlier Factor algorithm. Scoring: alert = min(1, max(0, kNN_distance_ratio_p95 - 1) / 2). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. structure.knn_distance_ratio_p95	3.1028	1.00	strong	Isolated spectrum	Rare cases	p95(mean kNN distance / median population mean kNN distance) in PCA-score space; not LOFalert = min(1, max(0, kNN_distance_ratio_p95 - 1) / 2)
Structure of the dataset	Isolation Forest score ? 95th percentile of the negated Isolation Forest score in retained PCA-score space. Larger raw scores indicate easier isolation under that model; the scale is implementation-dependent. The displayed alert reuses structure complexity rather than directly mapping this raw score. Scoring: alert follows structure complexity; raw score is implementation-dependent. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. structure.isolation_forest_score_p95	0.59045	1.00	strong	Atypical spectrum	Various causes	p95 IsolationForest anomaly score on PCA scoresalert follows structure complexity; raw score is implementation-dependent

PCA score plot — explore spectral variation

Source MicroNIR leaf spectra: 800 of 2,408 source observations are plotted. PC1 captures 72.1% and PC2 24.0% of the reported spectral variance: 96.1% together; 3.9% remains outside this view.

View plotted PCA coordinates

Plotted observation positions (not original sample IDs)
Point	PC1 score	PC2 score
1	-0.0070155	-0.0025276
2	-0.018959	-0.015931
3	-0.0052513	0.001531
4	-0.023318	-0.016738
5	-0.011695	0.0041054
6	-0.0053001	-0.0044631
7	-0.005858	0.0068397
8	-0.0084785	-0.0027737
9	-0.018439	-0.020294
10	-0.0058801	-0.0035172
11	-0.013858	-0.0075494
12	-0.0066294	-0.0047826
13	-0.0032485	0.00019517
14	-0.0013833	0.0021766
15	-0.017064	-0.0097195
16	-0.002297	-0.0046048
17	-0.0047747	0.00045147
18	-0.010278	-0.0013729
19	-0.0060608	0.0053991
20	-0.011526	-0.0088434
21	-0.0087464	-0.0013178
22	-0.0079155	-0.0023517
23	-0.0066722	0.0021483
24	-0.010594	-0.004417
25	-0.0024276	-0.0054129
26	-0.010615	-0.0017106
27	-0.0084568	0.0010552
28	-0.0038405	-0.0044606
29	-0.027934	-0.027697
30	-0.0064178	-0.0050254
31	0.0013399	-0.00032648
32	-0.003017	-0.0018628
33	-0.0063598	-0.0031437
34	-0.010723	-0.002156
35	-0.0067534	7.0538e-05
36	-0.012394	0.00040122
37	-0.0075623	-0.010823
38	-0.016723	-0.023679
39	-0.00067039	-0.0016262
40	-0.012309	0.0027482
41	-0.0078756	0.0032208
42	-0.010579	-0.00071818
43	-0.012978	-0.0085112
44	-0.011998	-0.0047626
45	-0.0026343	0.0009726
46	-0.013456	0.003701
47	-0.00036326	0.00063057
48	-0.012716	-0.0022168
49	-0.026346	-0.017795
50	0.00064461	-0.0018304
51	-0.0083461	-0.00068398
52	-0.0055974	-0.0019152
53	-0.013463	-0.0030443
54	-0.012824	-0.0073344
55	-0.011599	-0.00080194
56	-0.0050781	-7.9887e-05
57	0.0014897	-0.0098663
58	0.0036485	-0.0085325
59	-0.021398	-0.00776
60	-0.0046783	-0.00053671
61	-0.00030124	0.00030221
62	-0.015337	-0.00071791
63	-0.011551	0.00016557
64	-0.011606	-0.0053088
65	-0.0012594	0.001619
66	0.0018884	-0.0021553
67	-0.0088312	0.0037972
68	-0.0046464	-0.0053245
69	-0.010402	-0.0045169
70	-0.0055341	-0.0031114
71	-0.0039539	0.0022926
72	-0.0033947	-0.0011114
73	-0.0030023	0.0021389
74	-0.013669	0.003916
75	-0.0085074	0.0017985
76	-0.0009858	0.0012323
77	-0.016996	-0.010167
78	0.00050949	0.00021443
79	-0.007843	-0.0052843
80	-0.01504	0.0067686
81	-0.0089799	-0.0053101
82	-2.192e-05	0.0031928
83	-0.0078749	0.0013417
84	-0.024731	-0.019104
85	-0.00047755	-0.0021475
86	-0.011456	-0.0083621
87	-0.0065943	0.0013393
88	-0.010151	-0.0022943
89	0.001558	-0.0050437
90	-0.0032957	-0.0047779
91	0.0091506	-0.0073202
92	-0.012141	0.0053275
93	-0.015488	-0.0042781
94	0.00050366	-0.0010814
95	-0.0053528	0.0011916
96	-0.017074	-0.0053398
97	-0.0085056	0.005135
98	-0.0054538	-0.00091732
99	-0.0044974	-0.0021539
100	-0.019519	-0.0054451
101	-0.003156	-0.0035351
102	-0.002949	0.00076757
103	-0.013374	-0.0060987
104	-0.003433	-0.0082703
105	-0.0048643	0.002008
106	-0.011685	-0.0087767
107	-0.011355	0.0019279
108	-0.0007059	0.00010482
109	-0.0084309	-0.0066324
110	-0.01374	-0.0055592
111	-0.0047499	-0.0053702
112	-0.0097788	0.0057501
113	-0.012143	0.0043646
114	-0.010857	-0.0087688
115	4.7064e-06	-0.0036944
116	-0.0052201	0.0021812
117	-0.011106	0.001998
118	-0.014638	-0.0032024
119	-0.011654	-0.00078867
120	-0.0099271	0.0072225
121	-0.0050766	0.0029065
122	-0.010261	-0.0065194
123	-0.0078144	-0.005475
124	-0.0077719	0.0043744
125	-0.0055865	0.00064893
126	-0.0062937	0.0022748
127	-0.0028709	0.0020049
128	-0.0083493	-0.016115
129	-0.009753	-0.0027388
130	-0.007608	-0.0034642
131	-0.0041012	-0.0013745
132	-0.013572	0.0014314
133	-0.01146	-0.0034992
134	-0.015592	-0.0092727
135	-0.0036746	0.0062165
136	-0.0064102	-0.0032485
137	-0.013936	0.0011568
138	-0.0034844	-0.0023987
139	-0.0060554	-0.003345
140	-0.0095335	0.0037071
141	-0.008132	-0.0040741
142	0.0015304	0.00085998
143	-0.01066	-0.00065572
144	-0.0083074	-0.0033989
145	-0.0097606	-0.0033057
146	-0.0074083	-0.0026448
147	-0.010003	0.0017629
148	-0.010407	0.0010171
149	-0.010356	-0.0018427
150	-0.00039062	-0.007092
151	-0.0057879	0.00040418
152	-0.023022	-0.011907
153	-0.010024	-0.0015135
154	-0.011625	0.00011023
155	-0.010756	0.0017046
156	-0.011275	-9.5576e-05
157	-0.014638	0.0010154
158	-0.0060623	0.0058083
159	-0.0076806	-0.0019446
160	-0.0060804	-0.0024338
161	-0.0059543	0.0012526
162	-0.013521	-0.0047163
163	-0.0064194	-0.00042312
164	-0.010675	-1.2648e-05
165	-0.0076559	-0.0015681
166	-0.0054044	0.0058863
167	-0.0050612	-0.0033785
168	0.0018075	-0.0045638
169	-0.0070723	0.0062401
170	-0.0096299	-0.0030761
171	-0.0046028	0.0053365
172	-0.0097092	-0.012644
173	-0.01071	-0.0068638
174	-0.0091107	-0.00035888
175	-0.0089916	-0.00054522
176	0.001683	0.0010692
177	-0.0073947	-0.0039215
178	-0.012182	0.0029574
179	-0.0081853	-0.0073619
180	-0.011897	-0.0014185
181	-0.0085055	0.0023601
182	0.0020866	0.0018462
183	-0.013392	-0.00041855
184	-0.0064605	-0.0021744
185	-0.0078044	-0.0032312
186	-0.013642	-0.0096623
187	-0.022809	-0.0054752
188	-0.0062892	0.0063414
189	-0.0076849	-0.0015006
190	-0.004836	-0.0010347
191	-0.0097591	0.00026504
192	0.00054898	0.0026044
193	-0.021419	-0.0019588
194	-0.011019	0.0091835
195	-0.010753	0.0002896
196	-0.011067	0.0082233
197	-0.010995	-0.00075034
198	-0.0090497	-0.001185
199	-0.010347	-0.00074745
200	-0.0098735	0.0024507
201	-0.0049815	-0.0037183
202	-0.0064298	-0.0022663
203	-0.0098679	-0.0073932
204	-0.0083088	-0.0020799
205	-0.012307	-0.0083636
206	-0.0020992	0.0018054
207	-0.001353	0.0006119
208	-0.0072355	0.0052306
209	-0.015076	-0.0073123
210	-0.010641	-0.0087476
211	-0.0079642	0.0029989
212	-0.0030559	0.0035885
213	0.00027369	0.0033789
214	-0.012264	0.0014974
215	-0.0089881	-0.0029255
216	-0.012644	-0.006514
217	-0.00060883	0.0015721
218	-0.012562	0.0044552
219	-0.0067372	0.0015715
220	-0.010896	-0.006927
221	-0.0031572	5.8801e-05
222	-0.010686	-0.0021207
223	-0.011727	0.0047372
224	-0.017185	-0.016182
225	-0.0059011	-0.0014809
226	0.0024784	-0.00037978
227	-0.023382	-0.019054
228	-0.0096793	-0.0020344
229	-0.010987	0.00025744
230	-0.0065671	0.003322
231	-0.0058484	-0.0020699
232	-0.0083075	0.0011106
233	-0.024344	-0.020615
234	0.0011855	0.0017016
235	-0.0048869	-0.0033333
236	-0.0053213	-0.010006
237	-0.001828	-0.0015932
238	-0.0036093	-0.0026214
239	-0.015594	0.00098654
240	-0.010491	-0.00078142
241	-0.019302	-0.0016413
242	-0.013287	0.0012147
243	-0.018475	-0.0093249
244	0.0022685	0.0016301
245	-0.0077062	0.0019134
246	-0.0026252	-0.0034174
247	-0.0094012	-0.00055699
248	-0.010052	0.0042626
249	-0.016546	0.00092283
250	0.00043014	0.0014376
251	-0.02682	-0.02103
252	-0.011603	0.0036359
253	-0.014867	-0.00059842
254	-0.011005	-0.0012616
255	-0.0042847	0.0021458
256	-0.0070146	2.4908e-06
257	0.00025068	-0.010846
258	-0.0098218	0.0019181
259	-0.0091833	0.00031925
260	-0.0055348	-0.0040875
261	-0.0046294	-0.0011381
262	-0.013072	-0.006347
263	0.0032537	-0.0022465
264	-0.010783	-0.0032597
265	-0.010066	-0.00077878
266	-0.0068274	-0.00039395
267	-0.022643	-0.013814
268	-0.027692	-0.0088581
269	-0.0038219	0.0030976
270	-0.0034898	-0.0036547
271	-0.0071578	-0.0015882
272	-0.0019763	0.0023732
273	-0.011265	-0.0067005
274	-0.016218	-0.0079544
275	-0.0064441	0.0059553
276	0.0078277	-0.00092405
277	0.0040214	-0.006976
278	-0.0040206	-0.00082113
279	-0.0071113	0.0055086
280	-0.01682	-0.011475
281	0.0077685	-0.0064972
282	-0.0065725	-0.0032052
283	-0.010749	-0.0016425
284	-0.0045659	0.00059609
285	-0.0091648	0.0034474
286	-0.014767	0.003768
287	-0.0054544	0.00060136
288	-0.018414	0.004429
289	0.0019878	0.003955
290	0.0067712	0.0012635
291	0.0070653	-0.0014747
292	0.001763	-0.0012912
293	0.0078418	-0.0010148
294	0.00044732	0.0011095
295	0.012037	-0.0065482
296	0.013289	-0.0041715
297	0.01511	-0.0052245
298	-0.0092778	0.013214
299	0.0089802	-0.00027402
300	-0.00081133	0.0016639
301	0.013144	-0.0046886
302	0.00063793	0.0037105
303	0.0097262	-0.0034012
304	0.0080246	-0.0041966
305	0.0010576	0.0058773
306	0.012191	-0.0018338
307	0.0046447	0.0032995
308	0.0058106	0.0016004
309	0.0057273	-0.00010526
310	-0.0021048	0.0045154
311	0.0027527	0.0022624
312	0.011826	-0.00072146
313	0.0072152	0.000417
314	-0.0091236	0.013143
315	0.0087305	-0.0002934
316	0.0058533	0.0004592
317	0.0083134	-0.0052167
318	0.0078929	-0.0057453
319	-0.0029125	0.0049277
320	0.0047187	0.004199
321	0.0073982	0.001026
322	0.0097649	0.00039495
323	0.0066082	-0.0010499
324	0.0070272	0.0026169
325	0.0042097	0.0048729
326	0.0083311	0.0027826
327	0.0020695	0.0032569
328	0.0063362	0.0012429
329	0.001931	0.0045357
330	0.009811	-0.0030949
331	0.0089508	-0.0020846
332	0.0052975	0.0019904
333	0.0015112	0.0031711
334	0.002839	-0.0015355
335	-0.0015304	0.001648
336	0.0018485	0.0014557
337	0.0039063	0.0014152
338	0.010267	-0.0036345
339	0.010525	-0.0036999
340	0.015669	-0.0050881
341	0.0034276	0.0019801
342	0.010915	-0.002582
343	0.01327	-0.0022034
344	0.008915	-0.002927
345	-0.001473	0.0055389
346	0.0036718	0.0056083
347	0.020919	-0.0077159
348	0.0092294	-0.0021916
349	0.0045837	0.0010747
350	0.011309	-0.0039591
351	0.010546	-0.0016198
352	0.015917	-0.0016629
353	0.0049543	-0.0016705
354	0.008528	-0.0033203
355	0.0014282	0.0023557
356	0.012088	-0.0030196
357	0.0033141	0.0020072
358	0.0099251	-0.0022986
359	0.0053973	0.00047643
360	0.0041037	-0.0015361
361	0.0074742	-0.0022394
362	-0.0022737	0.004913
363	0.0048856	-0.0030007
364	0.016456	-0.0039712
365	0.0090349	0.00091127
366	-0.0088217	0.0047467
367	-0.003598	0.0017401
368	-0.0005921	-0.00080133
369	0.002265	0.0035557
370	0.0048612	0.002422
371	0.007055	-0.00027267
372	-0.00075322	0.0052557
373	-0.0033581	0.0087324
374	0.0076612	0.001274
375	0.0061681	0.0007189
376	-0.0027287	0.0064769
377	0.0069434	0.0011011
378	0.018154	-0.0062695
379	0.0058872	0.0025087
380	0.0064579	0.0015102
381	0.01177	-0.0058983
382	0.013475	-0.0055272
383	0.012933	-0.0064657
384	0.011623	-0.0046814
385	0.0060688	-0.00070835
386	0.0023165	-0.0024295
387	0.0046009	0.00034254
388	-0.0055279	0.011292
389	0.0095136	-0.0012545
390	-0.0012949	0.0043489
391	0.0086773	-0.0033946
392	0.0056087	-0.00044689
393	-0.012131	0.010151
394	-0.011757	0.0099481
395	-0.0058921	0.0094695
396	0.012053	-0.006156
397	-0.0016454	0.0017932
398	0.0021663	-0.0051329
399	-0.0017181	0.0051142
400	0.012483	-0.0034432
401	0.0033302	0.0047849
402	0.014948	-0.004035
403	0.0010173	0.0061872
404	0.0049199	-0.0018832
405	-0.0049336	0.0063847
406	0.0076457	-0.0032089
407	0.0023249	0.0026096
408	0.0073016	-0.00084835
409	0.0037558	0.0050245
410	-0.0028016	0.0084955
411	5.3982e-05	0.0042912
412	-0.002113	0.0063389
413	-0.0001596	0.0062495
414	0.0013674	0.0046049
415	-0.00031289	0.0057238
416	0.0072971	0.0012884
417	0.0059286	-0.0018878
418	0.0045017	0.0035631
419	0.00084276	0.0057147
420	-0.0030974	0.004714
421	0.0045057	0.0025984
422	0.013101	-0.0021121
423	0.0015404	0.0036909
424	0.0084158	-0.0033424
425	-0.0003672	0.0033904
426	0.0023881	0.00072896
427	-0.00010701	0.0033644
428	0.0031634	-6.9784e-05
429	-0.0022177	0.0032221
430	0.0062232	-0.00040271
431	-0.0031893	0.00453
432	0.003674	0.0055835
433	0.0024264	0.0035272
434	0.0096075	0.0028359
435	0.016258	-0.0055931
436	0.0047122	0.003725
437	0.016195	-0.0058038
438	0.0015804	-0.00044855
439	0.01641	-0.0068178
440	0.016814	-0.0070587
441	0.0021326	0.0018357
442	-0.0016501	0.003205
443	-0.0039266	0.0053541
444	-0.00018493	0.004372
445	0.0052274	0.0012176
446	0.0093934	-0.00094072
447	0.00067416	0.0061617
448	0.0015739	0.0046694
449	0.0014247	0.0059026
450	0.0085432	7.8935e-05
451	-0.0015487	0.0035525
452	0.0095391	-0.0050675
453	0.011293	-0.0042947
454	0.0045384	0.0025756
455	0.0055279	0.0004453
456	0.002344	0.0010806
457	0.0024507	0.0038121
458	0.018359	-0.0066207
459	0.0079704	0.00049958
460	0.0054366	0.00096589
461	0.0018852	0.0030569
462	0.0048957	0.0018974
463	0.01462	-0.0042619
464	-0.0059145	0.0095646
465	-0.010586	0.0054346
466	0.0042271	0.0049989
467	-0.010205	0.011393
468	0.0096255	-0.00062408
469	-0.0011824	0.0069401
470	0.012744	-0.0057611
471	-0.002254	0.0015024
472	0.0022246	0.0021147
473	0.0078881	-0.0012655
474	0.0033876	0.0014625
475	0.0061048	-0.0011385
476	0.016253	-0.0076148
477	0.010216	-0.0028018
478	0.010055	-0.00046708
479	-0.0062191	0.010968
480	-0.0046231	0.0091189
481	0.0018166	-0.00032812
482	-0.0075443	0.011465
483	-0.0062797	0.011082
484	0.0057041	-0.00013112
485	0.0033675	0.0009505
486	0.0019468	-0.00078048
487	0.0085019	-0.0026816
488	0.014061	-0.0067046
489	0.013655	-0.0057515
490	-0.00091927	0.0055079
491	-0.0040153	0.0077757
492	0.0025912	0.00535
493	0.017475	-0.0058481
494	0.0032264	0.00047976
495	0.0025039	0.0011135
496	0.0036444	0.000251
497	0.011439	-0.0056835
498	0.002604	-0.00042533
499	0.0075241	-0.0030505
500	-0.00069525	0.0049846
501	0.012607	-0.0027563
502	0.0060772	0.0014312
503	0.0032126	0.002659
504	-0.0011103	0.0074597
505	0.0041102	0.0034816
506	0.011131	-0.0018367
507	0.0092416	-0.0013446
508	0.0075539	-0.0021579
509	0.0070223	-0.0020877
510	0.0038545	-0.00071786
511	0.012909	-0.0056515
512	-0.00023824	0.0033112
513	0.0090856	-0.004381
514	0.0068575	0.00054115
515	0.011644	-0.0021014
516	-0.00050589	0.0055941
517	0.010684	-0.0032582
518	0.0018399	0.003516
519	0.0043718	0.005803
520	0.0087753	-0.0024905
521	-0.0054621	0.0024665
522	-0.0021476	0.00017909
523	0.0029038	0.003183
524	0.0011913	0.0085033
525	0.0053891	0.0012561
526	0.0096441	-0.001875
527	0.015256	-0.004912
528	-0.0030235	0.0047811
529	-0.0032873	0.0050341
530	0.0050728	-0.0049483
531	0.011357	-0.0045264
532	0.0088033	-0.0037948
533	0.0077621	-0.0053591
534	0.0019838	-0.0030528
535	0.00010314	0.0032069
536	0.0054348	0.0014785
537	0.01372	-0.0042945
538	0.0033018	0.002978
539	0.0068299	-0.00073391
540	0.0073648	-0.0027006
541	-0.0012153	0.0029055
542	0.0011656	0.0051639
543	0.010703	-0.0018846
544	0.0041232	0.0021107
545	0.007748	-0.0035177
546	0.0077253	-0.0055095
547	0.0026497	0.0028313
548	0.010178	-0.0014754
549	0.0072139	0.00046254
550	0.0053033	0.0030443
551	0.0053128	0.0032187
552	0.0035756	0.0039417
553	0.0017271	0.0054195
554	0.010403	-0.00041717
555	0.0020279	0.00232
556	-0.0016624	0.0045751
557	-0.0028151	0.00067958
558	0.010155	-0.0033301
559	-0.0014544	0.0032269
560	0.0050481	-0.0006975
561	0.020025	-0.0076061
562	0.0081963	0.0008224
563	0.013209	-0.0039986
564	0.003568	0.0041664
565	0.0072595	0.00023249
566	0.0012671	0.004942
567	0.0027486	0.0026772
568	0.00061535	0.0020392
569	0.013553	-0.0067561
570	0.0022241	0.0035466
571	0.0087073	-0.0017069
572	0.01628	-0.011493
573	0.001027	0.0068363
574	0.0032351	0.0035806
575	0.00037529	0.0023919
576	-0.0055796	0.0081862
577	-0.0021415	0.003948
578	-0.0046362	0.0056623
579	0.020954	-0.0073739
580	0.016237	-0.0036842
581	-0.0038867	0.0085487
582	0.0087174	-0.00050986
583	0.0029477	0.0029167
584	0.0019386	0.0035597
585	0.011442	-0.0027075
586	0.0096174	-0.0022655
587	0.0043128	0.0034793
588	0.012193	-0.0025424
589	0.011048	-0.0031843
590	0.0039687	0.0030341
591	0.025372	-0.011495
592	-0.0015409	0.0040602
593	-0.0076077	0.007885
594	0.0073884	0.0025595
595	0.010639	-0.0018048
596	0.0079592	0.0015767
597	-0.0005693	0.0052392
598	0.0033873	3.513e-05
599	0.0051314	-0.0019483
600	-0.0080769	0.0091106
601	-0.003764	0.0053767
602	0.0025586	0.00048127
603	0.0034771	0.0051484
604	-0.0012177	0.0086513
605	0.004097	0.0045271
606	0.011217	0.00035662
607	0.0053349	0.0041055
608	-0.00021039	0.0033796
609	-0.0054054	0.0064749
610	-0.0032662	0.0068233
611	0.0035053	0.0017093
612	0.0023682	0.0034842
613	-0.0012399	0.0075407
614	0.018388	-0.0059721
615	0.01301	-0.0031676
616	0.0047038	0.0010244
617	0.0085315	-0.0027598
618	0.0019738	0.0035506
619	0.0040359	0.0023635
620	0.013317	-0.0023989
621	0.011285	-0.0021823
622	-1.804e-05	0.0051647
623	0.0049721	-2.8383e-05
624	0.0045165	0.0015432
625	0.0077704	-0.00024318
626	-0.0022223	0.00675
627	-0.0026662	0.0077623
628	0.011577	-0.0013214
629	0.0057242	0.00024663
630	0.0063819	-0.00056749
631	0.010162	-0.002802
632	-0.0051748	0.0084909
633	0.00074297	0.0072051
634	-0.0074191	0.011404
635	-0.00011036	0.008042
636	0.0030792	0.0040301
637	0.0054531	0.0019902
638	-0.0053454	0.0029855
639	0.0037125	0.0038635
640	-0.0010962	0.0051386
641	0.011809	-0.0016599
642	0.0081358	-0.001677
643	-0.002152	0.0058551
644	0.0019635	0.0038516
645	0.0026236	0.0039543
646	0.0051632	0.0021049
647	0.0053034	0.0018991
648	0.0016492	0.0059007
649	-0.00476	0.0095763
650	0.0059513	0.0013415
651	-0.0060843	0.0070745
652	0.009739	-0.0029865
653	-0.0033541	0.0068843
654	0.01324	-0.0011645
655	0.015011	-0.0052501
656	0.015316	-0.0041289
657	0.007145	0.00097204
658	0.0070277	-0.0037895
659	0.0055428	-0.00052305
660	0.0062705	0.0018709
661	-0.0018838	0.00038609
662	0.0023606	-0.0028749
663	0.0016336	0.0035496
664	-0.0081265	0.010273
665	0.014544	-0.002989
666	0.007542	-0.00026696
667	0.0097362	0.00097143
668	0.0095443	0.00057903
669	0.0083716	-0.0046321
670	0.013916	-0.0069014
671	0.0062119	6.0701e-05
672	0.0011965	0.00311
673	-0.00040599	0.0071834
674	0.0028666	0.003259
675	0.0085167	-0.001485
676	0.0054263	-0.002225
677	0.0055763	-0.0039912
678	-0.0011772	0.0056863
679	-0.00054102	0.0059689
680	-0.00080749	0.0039679
681	-0.0027454	0.0075388
682	0.0054982	0.00022616
683	-0.0047149	0.0087499
684	0.0094383	-0.0070652
685	0.0011312	0.0047287
686	0.013095	-0.0019735
687	-0.0032124	0.0092152
688	-0.0012944	-0.00081973
689	-0.00016519	-0.0011763
690	0.0012573	0.005764
691	-0.00055544	0.0067656
692	-0.0082224	0.0065955
693	-0.0041979	0.008432
694	-0.00025396	0.0043127
695	0.0014716	0.0031329
696	-0.0023035	0.0057063
697	-0.0031965	0.0061781
698	0.0030139	0.0028377
699	-0.0098163	0.011295
700	0.010563	-0.0044484
701	0.0071571	-0.0016163
702	0.0084581	0.00093813
703	0.0024993	0.0022457
704	-0.0031063	0.0063029
705	0.0080936	-0.00050565
706	0.011181	-0.00052619
707	0.0026057	0.0044074
708	0.0024587	0.0053151
709	0.017092	-0.0053233
710	0.0097606	-0.00086491
711	0.0068832	7.4745e-05
712	0.0036415	0.0019575
713	0.0033836	0.0026962
714	0.015455	-0.0052987
715	0.014217	-0.00619
716	0.010328	-0.0023028
717	0.0038155	0.0034623
718	0.0091797	-0.001453
719	0.0085176	-0.00091989
720	0.013941	-0.0047121
721	0.01116	1.5655e-05
722	0.011109	-0.0023186
723	0.0052458	0.00044806
724	0.0013479	0.0032341
725	0.0055544	0.0012241
726	0.0071543	-0.0013247
727	-0.0029832	0.0051596
728	0.0083976	-0.0013616
729	0.0095386	-0.00097009
730	0.0026904	0.0028314
731	-0.003603	0.0070662
732	-0.0051151	0.0096881
733	0.0086421	0.00093849
734	0.0034087	0.0046305
735	0.015263	-0.0038149
736	0.0090807	0.0020522
737	0.00017441	0.0055005
738	0.011049	-0.0020622
739	0.0017476	0.0041035
740	0.01557	-0.0073249
741	0.011194	-0.0027759
742	0.0086106	0.00034564
743	0.004622	0.0041798
744	0.013757	-0.0037701
745	0.019192	-0.0062679
746	0.0058437	0.0017083
747	0.011217	-0.0026612
748	0.0067308	0.00029569
749	0.0034777	0.0027705
750	0.013767	-0.003381
751	0.015446	-0.0038884
752	-7.3426e-06	0.0068072
753	0.0018966	0.0011261
754	0.0066192	-0.0007173
755	0.0033154	-0.0025956
756	0.0075682	0.00017864
757	0.0051609	0.0012276
758	0.0019673	0.0028845
759	-5.5788e-06	0.0041139
760	0.0020841	0.0041713
761	0.00062251	0.0048365
762	0.011862	-0.0030187
763	0.001043	0.0035723
764	-0.00043047	0.0038931
765	0.0045146	0.0027611
766	-0.0032668	0.0096923
767	0.0025225	0.0035846
768	-0.0059162	0.0095951
769	-0.0040529	0.010302
770	-0.0032734	0.0024649
771	0.0039083	0.0031083
772	0.0023201	0.0032354
773	-0.00554	0.0081279
774	0.0069898	-0.0026735
775	0.0098013	-0.001612
776	0.0092857	-0.0034847
777	-0.0075313	0.0062826
778	-0.001905	0.0035031
779	0.0028553	0.0010857
780	0.010124	-0.0021255
781	0.0058594	0.0028612
782	0.012474	-0.001452
783	0.0101	-0.0017492
784	0.010372	-0.0012944
785	0.0038751	0.0018243
786	0.00029234	0.0013235
787	0.0097192	-0.0051669
788	-0.0029838	0.0085695
789	-0.0060431	0.01091
790	-0.0083111	0.012765
791	0.0066353	-0.0032107
792	0.0047024	-0.00048206
793	0.0015164	0.0018442
794	-0.0044272	0.0013598
795	0.0083079	-0.0038237
796	0.0051341	-0.001259
797	0.0091669	-0.0036714
798	-9.9408e-05	0.0019408
799	0.011989	-0.0036728
800	0.0083458	-0.00059956

PCA explained variance — how many dimensions?

View explained variance values

Variance per displayed principal component
Component	Spectral variance
PC1	71.25%
PC2	24.66%
PC3	1.81%
PC4	1.19%
PC5	0.25%
PC6	0.23%
PC7	0.21%
PC8	0.13%
PC9	0.08%
PC10	0.07%

X-Y spectral correlation 7

Spectral association with LMA

Read Pearson correlation (−1 to +1) along the spectral axis (nm); the solid curve shows |r| and the dashed curve shows signed r.

Spectral association with WCf

Read Pearson correlation (−1 to +1) along the spectral axis (nm); the solid curve shows |r| and the dashed curve shows signed r.

Spectral association with WQ

Read Pearson correlation (−1 to +1) along the spectral axis (nm); the solid curve shows |r| and the dashed curve shows signed r.

Target	max \|r\|	axis @ max	mean \|r\|	\|r\| ≥ .5
LMA	0.818	1.56e+03	0.509	58.4%
WCf	0.718	1.26e+03	0.21	12.0%
WQ	0.867	1.31e+03	0.641	75.2%
An	0.326	1.06e+03	0.177	0.0%
PSI	0.35	1.68e+03	0.167	0.0%
WUEintr	0.391	1.06e+03	0.239	0.0%
WUEinst	0.466	927	0.247	0.0%

NeoSpectra leaf spectra

X_neospectra · NIR · NeoSpectra

Read the signal and variation summaries for this source. Axis unit: nm; signal representation: not recorded.

Spectral signal summary

Sampling

How many observations and spectral channels were exported, and how the wavelength axis is spaced. Reading guide

Wavelengths ? Number of spectral channels. A channel can represent wavelength, wavenumber or another feature axis.	257
Axis range ? Lowest and highest recorded spectral-axis coordinates, in the stated unit.	1350–2550 nm
Mean spacing ? Average difference between adjacent spectral-axis coordinates; compare only within the same unit.	4.69 nm
Grid ? Uniform means adjacent axis spacings are approximately equal. An irregular grid may need care when applying derivatives.	irregular
Observations ? Exported spectral rows in this source, including repeated measurements when present.	2,116

Signal & quality

Signal level and roughness summaries. These depend on the signal representation and are not instrument acceptance tests. Reading guide

Value range ? Minimum and maximum recorded spectral values, in the original signal representation.	-0.0455 – 0.0119
Mean range ? Minimum and maximum of the mean spectrum across spectral channels.	-0.0387 – 0.0098
Mean level ? Average signal over observations and channels. The historical key mean_reflectance may also store other signal representations.	-0.002837
Area ? Trapezoidal integral of the mean spectrum over the sorted recorded axis; its unit is signal multiplied by axis unit.	-3.366
PTP ? Peak-to-peak: maximum minus minimum of the mean spectrum. Low values indicate little variation along the mean spectrum.	0.04849
Noise RMS ? Historical roughness sigma proxy from robust second differences. Narrow genuine bands can also raise it; this is not repeat-measured instrument noise.	2.3322e-05
Roughness S/N proxy ? Average absolute signal divided by the roughness proxy. Higher means smoother signal relative to its amplitude, not certified instrument SNR.	3.5e+02
Roughness S/N proxy dB ? 20 log10 of the roughness signal-to-noise proxy. Units are decibels; compare only compatible signal representations.	5e+01 dB
Dynamic range ? Maximum minus minimum of the mean spectrum. Its magnitude depends on the signal representation.	0.0485
Smoothness ? Mean per-spectrum RMS of second differences, in signal units. Smaller values indicate smoother spectra under this protocol.	0.0002793
Saturated ? Fraction at the observed global maximum; a heuristic indicator, not a comparison with known detector limits.	0.0%
X-outliers ? Number flagged by the qualification outlier method. An unusual spectrum can be valid and should not be discarded automatically.	265

Integrity & artefacts

Missing values, narrow spikes and abrupt jumps to investigate in the original spectra. Reading guide

NaN ratio ? Missing spectral cells divided by all spectral cells, displayed as a percentage.	0.00%
Inf count ? Number of positive or negative infinite spectral values; investigate invalid calculations or export.	0
Zero ratio ? Exactly zero spectral cells divided by all cells. Zero can be a valid measurement and is not automatically missing.	0.00%
Spike count ? Local narrow peaks flagged by the second-difference heuristic; real sharp spectral bands may also be flagged.	52,783
Spike rate ? Fraction of tested second-difference positions flagged as spikes, displayed as a percentage.	9.78%
Jump count ? Adjacent-channel changes flagged by the first-difference heuristic, potentially including detector joins.	12,194
Jump rate ? Fraction of tested adjacent-channel differences flagged as jumps, displayed as a percentage.	2.25%
Clip fraction ? Fraction of finite cells near the observed global minimum or maximum; inspect detector limits before concluding saturation.	0.00%

Shape & reference

Variation relative to this dataset's mean spectrum; the mean is not an independently measured calibration reference. Reading guide

Baseline slope ? Linear slope fitted to the mean spectrum over an axis normalized to 0–1. It describes total baseline change, not change per physical axis unit.	0.00094581
Curvature RMS ? Median per-spectrum RMS of second differences, summarizing curvature and high-frequency variation.	0.00027868
D1 RMS ? Median per-spectrum RMS of adjacent-channel first differences; reflects both real spectral structure and artefacts.	0.0013206
RMS to mean ? Root mean square distance to this dataset's mean spectrum, in signal units.	0.00060179
RMS p95 ? 95th percentile of distances to the mean spectrum: 95% of observations have a distance at or below this value.	0.001378
SAM to mean ? Spectral Angle Mapper: angle to the mean spectrum, in radians. Smaller angles indicate more similar shapes.	0.052196
SAM p95 ? 95th percentile of spectral angles to the dataset mean, in radians. It is less sensitive to overall amplitude than RMS.	0.12062
Affine offset p95 ? 95th percentile of absolute fitted offsets when matching each spectrum to the dataset mean.	0.00011831
Affine gain p95 Δ ? 95th percentile of the absolute gain deviation from 1 in the fit to the mean spectrum.	0.016701
Affine residual p95 ? 95th percentile of residual RMS after gain and offset correction to the mean spectrum.	0.0013731
Xcorr lag p95 ? 95th percentile of the absolute cross-correlation shift relative to the mean spectrum, in spectral channels.	0

Outliers & repeatability

Unusual spectra and within-ID dispersion. Repeated IDs do not by themselves establish technical repetitions. Reading guide

PCA Q p95/median ? Q (squared prediction error) is variation left outside the retained PCA subspace. The displayed ratio compares its 95th percentile with its median.	2.5
Hotelling T2 p95/median ? Hotelling T² measures distance within the retained PCA subspace, scaled by component variance. Ratio of the 95th percentile to the median.	2.8
Mahalanobis H p95/median ? Distance from the PCA-score center after variance scaling. Ratio of the 95th percentile to the median; large values identify relatively extreme spectra.	1.7
Repeat groups ? Number of IDs with several observations. Source documentation is needed to decide whether these are technical repeats, positions or time points.	0

Dimensionality (PCA)

Principal component analysis (PCA) compresses correlated wavelengths into components. This describes spectral variation, without using prediction targets. Reading guide

Effective rank ? Entropy-based summary of the variance distribution across components. Larger values indicate variation spread across more dimensions.	2.3
PCs → 95% var ? Number of principal components needed to explain 95% of spectral variance, when reached by the computed projection.	5
PCs → 99% var ? Number of principal components needed to explain 99% of spectral variance, when reached by the computed projection.	15
Top-10 cum. var ? Fraction of spectral variance explained by up to the first ten computed principal components, displayed as a percentage.	98.4%

Computed metric scores 30highest alert 1.00

Family	Computed metric	Value	Alert score	Heuristic level	Descriptive interpretation	Possible causes to verify	Calculation / scoring
Data integrity	Invalid/masked cell count (owner report) ? Count of cells reported unavailable by the native owner under its null or validity mask. This count is distinct from the historical NaN ratio; consult the recorded population and denominator. Scoring: No alert mapping is defined for this metric; an unavailable score does not mean a zero measured value. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.missing_cells	—	—	not calculable	Not enough information to score this metric on this dataset.	Missing data or validity mask; not evidence of an acquisition cause	Native Preflight count of cells unavailable under x/null and valid_mask; denominator stays in native_metric.denominators.n_cells. Not legacy NaN ratio.
Data integrity	NaN ratio ? Missing spectral cells divided by all spectral cells, displayed as a percentage. Scoring: alert = min(1, nan_ratio / 0.05). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.nan_ratio	0%	0.00	weak	Full spectrum	Acquisition/export error	count(isnan(X)) / X.sizealert = min(1, nan_ratio / 0.05)
Data integrity	Inf count ? Number of positive or negative infinite spectral values; investigate invalid calculations or export. Scoring: alert = min(1, inf_count / 1). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.inf_count	0	0.00	weak	Normal	Invalid calculations	count(isinf(X))alert = min(1, inf_count / 1)
Data integrity	Zero ratio ? Exactly zero spectral cells divided by all cells. Zero can be a valid measurement and is not automatically missing. Scoring: alert = min(1, zero_ratio / 0.05). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. integrity.zero_ratio	0%	0.00	weak	Normal	Export, saturation	count(X == 0) / count(finite X)alert = min(1, zero_ratio / 0.05)
Overall amplitude	Mean reflectance ? Average signal over observations and channels. The historical key mean_reflectance may also store other signal representations. Scoring: alert reuses baseline/shape drift because absolute reflectance ranges are technology-dependent. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.mean_reflectance	-0.0028366	0.34	weak	Too dark	Background, geometry	mean(X finite)alert reuses baseline/shape drift because absolute reflectance ranges are technology-dependent
Overall amplitude	Area under curve ? Trapezoidal integral of the mean spectrum over the sorted recorded axis; its unit is signal multiplied by axis unit. Scoring: alert reuses baseline/shape drift because area scale depends on axis and units. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.area_under_curve	-3.3659	0.34	weak	Normal	Probe distance	trapezoid(mean_spectrum, spectral_axis)alert reuses baseline/shape drift because area scale depends on axis and units
Overall amplitude	Peak-to-peak (PTP) ? Peak-to-peak: maximum minus minimum of the mean spectrum. Low values indicate little variation along the mean spectrum. Scoring: alert increases when dynamic range is abnormally flat. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.peak_to_peak	0.048494	0.00	weak	Strong variability	Saturation	max(mean_spectrum) - min(mean_spectrum)alert increases when dynamic range is abnormally flat
Overall amplitude	Variance ? Variance of all finite spectral cells, in squared signal units. It combines variation across observations and channels; a small value can describe a flat signal. Scoring: alert increases when variance/dynamic range is abnormally flat. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. amplitude.variance	0.00012196	0.00	weak	Normal or heterogeneous	Bad contact	var(X finite)alert increases when variance/dynamic range is abnormally flat
Noise	Roughness sigma proxy ? Historical roughness sigma proxy from robust second differences. Narrow genuine bands can also raise it; this is not repeat-measured instrument noise. Scoring: alert = noise_rms / signal_scale, saturated at 5%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. noise.roughness_sigma_proxy	2.3322e-05	0.01	weak	Stable	Lamp, detector	median MAD(second derivative) * 1.4826 / sqrt(6)alert = noise_rms / signal_scale, saturated at 5%
Noise	Roughness S/N proxy ? Average absolute signal divided by the roughness proxy. Higher means smoother signal relative to its amplitude, not certified instrument SNR. Scoring: alert decreases with SNR dB; >=40 dB is treated as low alert. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. noise.roughness_signal_to_noise_proxy	349.39	0.00	weak	Good signal	Acquisition	mean(abs(X)) / noise_rmsalert decreases with SNR dB; >=40 dB is treated as low alert
Noise	Bandwise roughness S/N proxy ? Minimum channel-wise ratio of absolute mean signal to local second-difference roughness. A smaller ratio identifies a weak or rough spectral region; genuine narrow bands and near-zero signal can also lower it. This is not certified instrument SNR. Scoring: alert decreases with worst-band SNR dB; >=35 dB is treated as low alert. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. noise.bandwise_roughness_signal_to_noise_proxy_min	1.7076	0.87	strong	Problem area	Detector	min(abs(mean_spectrum) / local second-derivative noise)alert decreases with worst-band SNR dB; >=35 dB is treated as low alert
Local artifacts	Spike count ? Local narrow peaks flagged by the second-difference heuristic; real sharp spectral bands may also be flagged. Scoring: alert follows spike_rate, saturated at 1%. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.spike_count	52,783	1.00	strong	Artifacts	Cosmic rays, splice	count robust outliers in second derivativealert follows spike_rate, saturated at 1%
Local artifacts	Spike rate ? Fraction of tested second-difference positions flagged as spikes, displayed as a percentage. Scoring: alert = min(1, spike_rate / 0.01). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.spike_rate	9.78%	1.00	strong	Suspicious spectrum	Interpolation	spike_count / (n_samples * (n_features - 2))alert = min(1, spike_rate / 0.01)
Local artifacts	Jump count ? Adjacent-channel changes flagged by the first-difference heuristic, potentially including detector joins. Scoring: alert follows jump_rate, saturated at 1%. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.jump_count	12,194	1.00	strong	Detector connection	Splice	count robust outliers in first derivativealert follows jump_rate, saturated at 1%
Local artifacts	Jump rate ? Fraction of tested adjacent-channel differences flagged as jumps, displayed as a percentage. Scoring: alert = min(1, jump_rate / 0.01). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.jump_rate	2.25%	1.00	strong	Spectral problem	Calibration	jump_count / (n_samples * (n_features - 1))alert = min(1, jump_rate / 0.01)
Local artifacts	Clip fraction ? Fraction of finite cells near the observed global minimum or maximum; inspect detector limits before concluding saturation. Scoring: alert = min(1, clip_fraction / 0.01). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. artefacts.clip_fraction	0.000368%	0.00	weak	Normal	Detector saturated	fraction of finite cells equal to repeated min/max extremaalert = min(1, clip_fraction / 0.01)
Spectral form	Baseline slope ? Linear slope fitted to the mean spectrum over an axis normalized to 0–1. It describes total baseline change, not change per physical axis unit. Scoring: alert = abs(slope / signal_scale), saturated at 0.5. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. shape.baseline_slope	0.00094581	0.04	weak	Stable	Illumination	linear slope of mean_spectrum over normalized axisalert = abs(slope / signal_scale), saturated at 0.5
Spectral form	Curvature RMS ? Median per-spectrum RMS of second differences, summarizing curvature and high-frequency variation. Scoring: alert = curvature_rms / signal_scale, saturated at 1%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. shape.curvature_rms	0.00027868	0.57	medium	Unusual shape	Background, splice	median RMS(second derivative per spectrum)alert = curvature_rms / signal_scale, saturated at 1%
Spectral form	D1 RMS ? Median per-spectrum RMS of adjacent-channel first differences; reflects both real spectral structure and artefacts. Scoring: alert = d1_rms / signal_scale, saturated at 5%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. shape.d1_rms	0.0013206	0.54	medium	Structured spectrum	Biology or artifact	median RMS(first derivative per spectrum)alert = d1_rms / signal_scale, saturated at 5%
Multivariate outliers	PCA Q (SPE) ? Q (squared prediction error) is variation left outside the retained PCA subspace. The displayed ratio compares its 95th percentile with its median. Scoring: alert = min(1, pca_q_ratio / 8). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. outliers.pca_q_ratio	2.4967	0.31	weak	Compliant	Artifact, mixture	p95(Q/SPE residual) / median(Q/SPE residual)alert = min(1, pca_q_ratio / 8)
Multivariate outliers	Hotelling T² ? Hotelling T² measures distance within the retained PCA subspace, scaled by component variance. Ratio of the 95th percentile to the median. Scoring: alert = min(1, hotelling_t2_ratio / 8). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. outliers.hotelling_t2_ratio	2.8399	0.35	weak	Central	Natural variability	p95(Hotelling T2) / median(Hotelling T2)alert = min(1, hotelling_t2_ratio / 8)
Multivariate outliers	Mahalanobis H ? Distance from the PCA-score center after variance scaling. Ratio of the 95th percentile to the median; large values identify relatively extreme spectra. Scoring: alert = min(1, mahalanobis_h_ratio / 4). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. outliers.mahalanobis_h_ratio	1.6852	0.42	medium	Global outlier	Different domain	p95(sqrt(T2)) / median(sqrt(T2))alert = min(1, mahalanobis_h_ratio / 4)
Comparison to average spectrum	RMS to mean spectrum ? 95th percentile of distances to the mean spectrum: 95% of observations have a distance at or below this value. Scoring: alert = RMS_p95 / signal_scale, saturated at 25%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. reference.rms_to_mean_spectrum_p95	0.001378	0.11	weak	Typical	Domain shift	p95 RMS distance to dataset mean spectrumalert = RMS_p95 / signal_scale, saturated at 25%
Comparison to average spectrum	Spectral Angle Mapper (SAM) ? 95th percentile of spectral angles to the dataset mean, in radians. It is less sensitive to overall amplitude than RMS. Scoring: alert = min(1, SAM_p95 / 0.35 rad). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. reference.sam_to_mean_spectrum_p95	0.12062	0.34	weak	Similar	Background, geometry	p95 spectral angle to dataset mean spectrumalert = min(1, SAM_p95 / 0.35 rad)
Repeatability	RMS intra-ID ? Within-ID root mean square deviation from the group's mean spectrum, in signal units. Repeated IDs require documentation before this spread can be called technical repeatability. Scoring: alert = RMS_intra_ID / signal_scale, saturated at 10%. Signal scale is the larger of the absolute mean level and absolute peak-to-peak range, with a floor of 1e-12. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. repeatability.rms_intra_id	—	—	not calculable	Not enough information to score this metric on this dataset.	Positioning	median RMS distance to repeated-sample centroidalert = RMS_intra_ID / signal_scale, saturated at 10%
Repeatability	SAM intra-ID ? Within-ID spectral angle to the group's mean spectrum, in radians. Repeated IDs require documentation before this spread can be called technical repeatability. Scoring: alert = min(1, SAM_intra_ID / 0.15 rad). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. repeatability.sam_intra_id	—	—	not calculable	Not enough information to score this metric on this dataset.	Acquisition	median SAM to repeated-sample centroidalert = min(1, SAM_intra_ID / 0.15 rad)
Repeatability	CV intra-ID ? Within-ID coefficient of variation: standard deviation relative to absolute mean; unstable near zero signal. Repeated IDs require documentation before this spread can be called technical repeatability. Scoring: alert = min(1, CV_intra_ID / 0.25). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. repeatability.cv_intra_id	—	—	not calculable	Not enough information to score this metric on this dataset.	Operator	median within-ID band CValert = min(1, CV_intra_ID / 0.25)
Structure of the dataset	PCA score density ? Inverse median distance to nearby observations in retained PCA-score space. Larger raw values mean more closely spaced observations; scale depends on the PCA representation. The alert uses density variation, so raw density alone does not establish an anomaly. Scoring: alert follows density_cv/profile structure complexity, not raw density alone. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. structure.pca_score_density	380.66	0.58	medium	Subpopulations	Different batches	1 / median kNN distance in PCA score spacealert follows density_cv/profile structure complexity, not raw density alone
Structure of the dataset	kNN distance ratio (not LOF) ? 95th percentile of mean k-nearest-neighbor distance divided by the population median of that distance, in PCA-score space. Values above 1 indicate an upper tail with more isolated observations; valid rare spectra can also be isolated. This distance ratio is not the Local Outlier Factor algorithm. Scoring: alert = min(1, max(0, kNN_distance_ratio_p95 - 1) / 2). Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. structure.knn_distance_ratio_p95	1.7312	0.37	weak	Normal population	Rare cases	p95(mean kNN distance / median population mean kNN distance) in PCA-score space; not LOFalert = min(1, max(0, kNN_distance_ratio_p95 - 1) / 2)
Structure of the dataset	Isolation Forest score ? 95th percentile of the negated Isolation Forest score in retained PCA-score space. Larger raw scores indicate easier isolation under that model; the scale is implementation-dependent. The displayed alert reuses structure complexity rather than directly mapping this raw score. Scoring: alert follows structure complexity; raw score is implementation-dependent. Higher alert scores indicate stronger heuristic concern, not a fault probability or an acceptance decision. A dash means unavailable. structure.isolation_forest_score_p95	0.53835	0.58	medium	Atypical spectrum	Various causes	p95 IsolationForest anomaly score on PCA scoresalert follows structure complexity; raw score is implementation-dependent

PCA score plot — explore spectral variation

Source NeoSpectra leaf spectra: 800 of 2,116 source observations are plotted. PC1 captures 81.5% and PC2 8.1% of the reported spectral variance: 89.5% together; 10.5% remains outside this view.

View plotted PCA coordinates

Plotted observation positions (not original sample IDs)
Point	PC1 score	PC2 score
1	0.012836	-0.0039602
2	0.0095191	-0.004705
3	0.010544	-0.0056602
4	-0.00017074	-0.005335
5	0.0057352	-0.0064465
6	0.019414	-0.0031732
7	0.0030046	-0.0063713
8	0.010838	-0.0027409
9	0.0065515	-0.0056917
10	0.0025203	0.00042725
11	0.0022885	-0.0020272
12	0.020971	-0.0026906
13	0.0013232	-0.0037631
14	0.0067181	-0.0055105
15	0.0072091	0.00021271
16	-0.0015108	-0.0028972
17	0.0109	-0.0031715
18	0.0022309	-0.0028036
19	0.003723	-0.0043549
20	0.0090598	-0.0028022
21	0.0033673	-0.0016123
22	0.01407	-0.0037555
23	0.010865	-0.003302
24	0.0014956	-0.0048709
25	0.0038416	-0.0035004
26	-0.0073893	-0.0066603
27	-0.0019749	-0.0038892
28	1.7641e-05	0.00028796
29	0.010178	-0.0021534
30	0.020343	-0.00064743
31	0.0091787	-0.00066027
32	0.011907	-0.0026268
33	0.011004	-0.0015906
34	0.016439	-0.0029818
35	0.0035602	-0.0032234
36	0.010816	-0.002193
37	0.014844	-0.0045843
38	0.0051183	-0.0053763
39	0.001368	0.0010834
40	0.0119	-0.0018036
41	0.010845	-0.0014354
42	0.012381	-0.0051486
43	0.0089244	-0.00066767
44	0.0084418	-0.0022553
45	0.016656	-0.00093052
46	-0.0080563	-0.0056944
47	0.0028215	-0.0054776
48	0.0067011	-0.0035163
49	0.016012	-0.0035155
50	0.013218	-0.00055853
51	0.0062524	-0.0043166
52	0.011361	-0.0019624
53	0.013408	-0.0062158
54	0.006555	-0.0040497
55	-0.0039429	-0.0042009
56	0.012981	-0.0036077
57	0.0072344	-0.0023996
58	0.011854	-0.0053516
59	0.015842	-0.0036559
60	0.011627	-0.001833
61	0.0061613	-0.0046454
62	0.0061104	-0.0051657
63	0.01601	-0.00089818
64	0.011547	-0.0033711
65	0.0098038	-0.0013241
66	0.0026919	-0.0047523
67	0.002497	-0.0038203
68	0.0079165	-0.001915
69	0.014252	-0.00054699
70	0.0082807	-0.0063296
71	0.015366	-0.0011904
72	0.0083095	-0.0059816
73	0.016682	-0.0056753
74	0.0055178	-0.0020058
75	0.0098997	-0.0024115
76	0.0086508	-0.003981
77	0.010662	-0.0012971
78	0.0085508	-0.00044857
79	-0.0012542	-0.0056512
80	0.0026825	-0.0030707
81	0.0085617	-0.0040848
82	0.009773	-0.0018056
83	0.014256	-0.0016912
84	0.013549	-0.0011816
85	0.01398	-0.0021644
86	0.0069173	-0.003726
87	0.010072	-0.00043764
88	0.011651	-0.0018909
89	0.017393	-0.0034679
90	-0.0075191	-0.0090926
91	0.0088842	-0.0038414
92	0.0083383	-0.0011051
93	0.018129	-0.0023643
94	0.019871	-0.0029651
95	0.0095131	-0.00054771
96	0.012346	-0.0016478
97	0.0067061	-0.003965
98	0.0089409	-0.0052696
99	-2.4705e-05	-0.0058073
100	0.0037882	-0.0055854
101	0.021234	-0.0019146
102	0.003443	-0.0063538
103	0.025233	-0.0017136
104	0.014949	-0.0021728
105	-0.0061615	-0.0051705
106	0.010053	-0.0034119
107	0.013361	-0.0011434
108	-0.00050884	-0.0057587
109	-8.2205e-05	-0.0022779
110	0.011191	-0.0039904
111	0.016056	-0.00049761
112	-0.00083076	-0.0050044
113	0.011849	4.7777e-05
114	0.01216	-0.0060249
115	0.0064496	-0.0055765
116	0.0095642	-0.006196
117	0.0097183	-0.0040106
118	0.0088296	-0.0014922
119	-0.0033754	-0.0060699
120	0.0042457	-0.00306
121	0.0066825	-0.0033165
122	-0.0023512	-0.0070342
123	0.0055563	-0.0026169
124	0.012004	-0.0021473
125	0.0085307	-0.0034794
126	0.0031496	-0.002999
127	0.0091064	-0.0038582
128	0.016209	-0.0028998
129	0.016565	-0.0020924
130	0.0040114	-0.00277
131	0.0050565	-0.0047814
132	0.018091	-0.0016488
133	0.0075703	-0.0027558
134	0.011396	-0.0019911
135	0.01182	-0.001344
136	0.0099619	-0.0013805
137	0.012688	-0.0029699
138	0.0166	-0.0012423
139	0.011037	0.00036418
140	0.012436	-0.002127
141	0.010148	-0.0034624
142	0.0090038	0.0033086
143	0.0092593	-0.0034257
144	0.025855	-0.0040593
145	0.0071296	-0.0016945
146	0.016822	-0.0023892
147	0.0065347	-0.0041563
148	0.015878	-0.0048507
149	0.010056	-0.0013038
150	0.010716	-0.0038198
151	-0.0080426	-0.0053922
152	0.007728	-0.0070057
153	0.0094348	-0.0035791
154	0.01238	-0.0038015
155	0.010922	-0.0020066
156	0.010831	-0.0026893
157	0.00080542	-0.0049888
158	0.0076977	-0.0064601
159	0.0059845	-0.00087943
160	0.010838	-0.0016212
161	0.015611	-0.0015495
162	0.0030853	0.006733
163	0.0040135	0.00030132
164	0.022448	0.00012692
165	0.014774	-0.0018898
166	0.011751	-0.0027403
167	0.017909	-0.0019009
168	0.018794	-0.0027974
169	0.021119	-0.0023156
170	0.0030688	-0.0021722
171	0.019894	-0.0016778
172	0.016919	-0.0026146
173	0.0031616	-0.0040443
174	0.019841	-0.00046905
175	0.011533	-0.0022515
176	0.0038747	-0.0053477
177	0.002476	-0.0015802
178	0.012386	-0.0031822
179	-0.0030131	-0.0017331
180	0.011926	-0.0032266
181	0.0095411	-0.0052677
182	0.023728	-0.0017642
183	0.019207	-0.0027226
184	0.0096103	-0.00022866
185	-0.0024484	-0.0047383
186	-0.0027967	-0.0075603
187	0.010805	-0.0019721
188	0.021077	-0.003293
189	-0.0021499	-0.0041891
190	0.007713	-0.0048752
191	0.027164	-0.0026114
192	0.0071573	-0.0028777
193	0.0055666	-0.0047973
194	0.015133	-0.0047637
195	0.0011152	-0.0047474
196	0.015409	-0.0045335
197	0.0056569	-0.005953
198	0.012568	-0.0045635
199	0.021781	-0.0032828
200	0.0091378	-0.002428
201	0.0036785	-0.0040546
202	-0.00034387	-0.003051
203	0.010194	-0.0014781
204	0.0053267	-0.0024845
205	0.0074842	-0.0024857
206	0.012022	-0.0029873
207	0.0055909	-0.0038255
208	0.0022502	-0.0062579
209	0.0081282	-0.0020016
210	0.011927	-0.0051447
211	-0.011884	-0.0075792
212	0.0054029	-0.0051221
213	0.0049483	-0.0089179
214	0.0073057	-0.0013337
215	0.010489	6.1677e-05
216	-0.013518	-0.0047974
217	-0.0078413	-0.0032354
218	0.00038736	-0.0026862
219	0.012758	0.0029867
220	0.010915	0.0002632
221	-0.0070583	-0.00028349
222	-0.0075168	0.0018133
223	-0.0031176	0.0014374
224	-0.007781	0.00085234
225	-0.011685	0.0058855
226	-0.011685	0.016481
227	-0.010275	0.011793
228	0.014356	0.01255
229	0.002085	0.0037045
230	0.0013673	0.00077735
231	0.0033202	0.002687
232	0.0056477	0.0024612
233	0.002251	0.0030301
234	0.0025517	-0.0003185
235	-0.022289	-0.00040702
236	-0.0012189	-0.0015041
237	-0.021094	-0.0032253
238	-0.0011446	0.0054278
239	-0.00805	-0.0022889
240	0.015001	0.0032464
241	0.0085843	0.0049044
242	0.0010427	0.0011773
243	0.0059548	0.0035477
244	-0.012001	0.00069138
245	-0.0055654	0.0053254
246	0.028243	-0.00081003
247	-0.00086443	0.0027599
248	-0.0097543	-7.8244e-05
249	-0.0089961	0.0010548
250	-0.0017895	0.0034406
251	0.00086034	0.0047242
252	-0.013305	-0.0014479
253	0.014623	-0.0010244
254	0.012067	0.0056878
255	-0.0027106	0.00012398
256	-0.0029647	-0.0013358
257	-0.014142	-0.0018636
258	-0.018104	0.0063496
259	0.0054127	-4.5595e-05
260	0.0075912	0.0024316
261	-0.014701	-0.00077737
262	0.0014469	0.0039905
263	-0.021981	-0.0025194
264	0.0058298	0.0040214
265	0.0037515	0.00045411
266	-0.016366	-0.00080465
267	-0.014687	-0.00075671
268	-0.020828	-0.0023221
269	-0.01049	-0.0018083
270	-0.011074	0.0005657
271	-0.00021248	0.00052151
272	0.0078671	0.0016015
273	-0.015062	-0.003835
274	-0.0014284	0.00088695
275	-0.00032663	0.0028736
276	-0.026553	-0.0024932
277	-0.0066664	0.0011709
278	-0.0081872	0.0009387
279	-0.0082656	-0.00043015
280	-0.010306	-0.0007551
281	-0.016219	-0.0018267
282	0.0012411	0.0021205
283	-0.016657	0.00078765
284	-0.022071	-0.0035775
285	-0.023467	-0.0012738
286	-0.01175	0.0043426
287	-0.014056	-0.0038851
288	-0.020137	-0.00616
289	-0.014587	0.0052335
290	-0.019884	0.0036558
291	-0.0085295	-0.0021635
292	0.003067	-0.0022873
293	-0.0079337	0.0040583
294	-0.016239	-8.3635e-05
295	0.0030438	0.00016645
296	-0.0067104	-0.0023422
297	-0.0022936	-0.0025153
298	0.0077609	0.011674
299	-0.0050132	0.0016011
300	-0.0025898	0.003389
301	0.0026664	0.00060136
302	0.0072114	-0.00078809
303	-0.0046614	0.0043886
304	0.0051368	0.00042466
305	0.0045505	0.0048621
306	0.012737	0.0047768
307	0.00037578	0.0010789
308	-0.0014587	-0.001763
309	-0.011268	0.0017661
310	-0.009824	-0.00053942
311	-0.002337	0.0015244
312	0.015198	0.0026478
313	-0.0023546	0.0017775
314	-0.0014855	0.00067776
315	-0.0047676	0.0010553
316	-0.019255	-0.00015641
317	-0.019269	-0.0074528
318	0.0001734	0.0026776
319	-0.0037729	0.0075654
320	-0.0018142	0.0055381
321	0.0027553	0.0025731
322	-0.012743	0.00061993
323	-0.0029805	0.0049387
324	-0.0077785	0.0012877
325	0.0035736	0.0047671
326	-0.0083277	0.0018892
327	-0.022758	-0.0033627
328	-0.0030835	-0.003277
329	-0.00067274	-0.0011623
330	0.020253	0.0046196
331	-0.011092	-0.0032375
332	0.0065614	0.0017205
333	-0.021521	-0.0053545
334	0.0077265	0.0033295
335	-0.013874	0.00068545
336	-0.017936	0.00079515
337	-0.0017439	0.0037469
338	-0.0041708	-0.0011286
339	-0.011256	0.0017418
340	-0.0087561	-0.0028903
341	-0.010565	-0.0012368
342	-0.0040994	0.0022434
343	-0.020628	0.00058252
344	0.010542	0.0057743
345	-0.014403	-0.0012203
346	-0.0057114	-0.0011371
347	-0.0094825	-0.0026388
348	0.0072385	-0.0030501
349	-0.0045127	-0.0022688
350	0.0054072	0.0011582
351	0.018465	0.0036733
352	-0.027699	-0.00087564
353	0.0053285	0.00098035
354	-0.017578	-0.0022719
355	-0.0057466	-0.00069509
356	-0.0015292	0.00046491
357	0.027684	0.0016389
358	0.008252	0.00043988
359	-0.0015678	-0.0015211
360	0.0067825	0.0056767
361	-0.00038057	-0.0021596
362	0.01259	-0.0013775
363	0.0085263	0.0035605
364	0.0022573	0.0061365
365	0.0032501	-0.0054675
366	0.00022508	0.0030884
367	0.0026216	-0.0023231
368	0.0035363	0.0014306
369	-0.012227	0.00022388
370	-0.012054	-0.00030469
371	-0.017413	0.00073937
372	-0.013439	0.00049362
373	0.0076431	0.00019838
374	0.0079496	-0.0011867
375	0.014631	0.0023393
376	0.0014278	0.00066439
377	-0.0055647	-0.0033984
378	0.0010715	-0.0042814
379	0.0033536	0.0021843
380	-0.00049583	-0.0011748
381	-0.0057851	-0.00093451
382	-6.9821e-05	0.0012868
383	-0.0052259	-0.0046975
384	-0.0092886	-0.0049447
385	0.0017282	0.00010489
386	-0.025193	-0.0059719
387	-0.0061523	-0.0021845
388	-0.032997	-0.0015567
389	-0.017107	0.00011226
390	-0.0041252	0.0005439
391	0.0030343	0.003658
392	-0.019326	-0.0020261
393	-0.0067243	0.0033969
394	-0.00641	0.0031797
395	0.0027647	-0.0013712
396	-0.013397	-0.0025967
397	-0.0025031	-0.001294
398	-0.003579	0.0023035
399	-0.0049848	0.0014346
400	0.0069569	0.0032776
401	0.004599	-0.0010482
402	-0.01115	0.0015324
403	-0.0024805	0.0037979
404	-0.0099368	0.00067537
405	-0.006212	0.0011142
406	-0.014766	-0.0029515
407	-0.012257	-0.0025536
408	0.015493	-0.0020193
409	-0.0015425	0.0015204
410	-0.00013692	-0.00081991
411	-0.013946	-0.0037949
412	-0.010539	0.0058309
413	-0.02094	0.00053766
414	-0.0046604	-0.0033517
415	-0.0019547	0.0013509
416	0.0066866	0.0071613
417	-0.0037561	0.0044854
418	0.00044903	-0.0014644
419	0.00049961	0.00067249
420	-0.024449	0.00014269
421	-0.021937	-0.0019122
422	-0.0045115	-0.0032386
423	-0.018985	-0.0040012
424	-0.012076	0.0022847
425	0.0092778	0.0044377
426	-0.015455	0.00058339
427	-0.009031	-0.0011607
428	-0.0027498	0.0016391
429	-0.017697	-0.0035957
430	-0.0015992	0.00055881
431	-0.0077438	-0.00048582
432	-0.017759	-8.2754e-05
433	-0.0097459	0.00094538
434	0.023804	0.0053544
435	0.026685	0.0045202
436	0.0041516	0.0085279
437	0.00041907	0.0017525
438	0.00087738	0.0030621
439	-0.0049909	0.0017374
440	0.012184	-0.0056255
441	0.0078853	0.0037895
442	-0.015019	-0.0053131
443	-0.016014	-0.0039512
444	0.016852	-0.00096552
445	-0.021192	-0.004152
446	-0.018476	-0.0047248
447	0.011336	0.0058859
448	0.0082009	0.0052262
449	-0.0030004	0.0019191
450	0.014415	0.0035262
451	-0.013261	-0.0015385
452	0.0080458	0.00088365
453	0.0030097	0.0058525
454	-0.00013949	0.0017806
455	-0.00064963	0.00090583
456	-0.01888	-0.0043214
457	0.0081193	0.0080728
458	-0.0148	0.00076944
459	-0.01194	0.0023563
460	-0.0031856	0.00020064
461	-0.0034213	0.0060642
462	-0.016232	0.0036515
463	0.0028344	0.0073776
464	0.0020311	0.0084774
465	-0.012611	-0.0011798
466	0.019583	0.0041156
467	0.012051	0.0067814
468	0.0067134	0.0031898
469	-0.0015362	0.0050601
470	-0.017567	0.00026153
471	0.0014361	0.0007715
472	-0.011404	0.0015842
473	-0.010012	0.0011113
474	-0.006624	0.00068368
475	-0.0019739	-7.7302e-05
476	0.015225	0.0015903
477	-0.0019929	-0.00016987
478	0.0044198	0.0027251
479	-0.0020425	0.0059932
480	-0.015194	0.0023504
481	-0.0016289	0.005899
482	0.0060502	0.0056569
483	0.0025107	0.0027957
484	0.0069357	0.0033855
485	-0.01174	0.00060554
486	0.0046541	0.0035191
487	0.0080988	0.0029334
488	-0.0009611	0.0067796
489	0.0057818	0.004579
490	-0.0084567	0.0029489
491	-0.0034961	0.0023317
492	-0.015175	0.0022844
493	0.0045571	-0.00013991
494	0.013345	0.0010684
495	0.0070137	-0.0019842
496	0.0070466	-0.00088413
497	-0.0073965	0.0054993
498	-0.0032973	-0.0002503
499	0.0072726	0.00087165
500	0.0040404	-0.00070249
501	0.0056541	0.0033352
502	-0.013389	-0.00032855
503	0.00096277	0.0078713
504	-0.010981	0.0021601
505	-0.0094929	-0.00087534
506	-0.012208	-0.0018623
507	-0.01171	-0.00023612
508	-0.01077	0.0020589
509	0.0027715	0.0069169
510	0.0095038	0.0050695
511	0.0013958	0.0057358
512	-0.014229	0.0031051
513	0.0070882	0.0071456
514	0.0068124	0.0048633
515	-0.014585	0.00040184
516	-0.023905	0.001167
517	-0.0010086	-0.00018359
518	0.005543	0.0060372
519	-0.0061045	0.0060984
520	0.013312	0.00043351
521	-0.0020125	0.0057542
522	-0.0035004	0.00458
523	0.00083826	0.0027338
524	0.00060552	0.0057614
525	-0.0007934	0.0033336
526	-0.0031562	-0.00015058
527	-0.0090606	-0.003464
528	-0.00019668	-0.0023124
529	-0.0071712	-0.0031492
530	-0.0086869	0.00018649
531	-0.0021785	0.0051305
532	-0.0086677	0.00072462
533	-0.020941	-0.0038599
534	-0.011157	-0.0036419
535	0.011993	0.0019556
536	-0.0069897	-0.0013746
537	0.00048204	0.00062768
538	-0.00042736	0.00099292
539	-0.03003	-0.0028294
540	-0.011108	0.0034651
541	0.0017119	-0.0035434
542	0.014331	-0.00070571
543	0.0050087	0.0016932
544	0.010583	-0.0022788
545	0.0030828	0.0040096
546	0.021088	-0.0013823
547	-0.024896	-0.0010097
548	-0.015287	0.0030712
549	-0.030988	-0.0024468
550	-0.0018001	0.0055471
551	0.0031488	0.0048301
552	-0.0064688	0.0021061
553	-0.01411	-0.00039625
554	-0.0011855	-6.4854e-05
555	0.00069259	0.0027142
556	0.0030088	-0.001719
557	-0.015395	-0.00056583
558	-0.0046464	0.0017682
559	0.0033899	0.0029626
560	-0.0094425	0.0063769
561	0.00078793	0.00011029
562	0.0021433	0.00087829
563	0.010768	0.0017611
564	0.0054841	0.00069069
565	-0.0012679	0.0033503
566	0.0075262	-0.004731
567	-0.0037326	-0.002711
568	-0.014581	-0.0020478
569	0.0054868	0.0020645
570	-0.013101	-0.0007394
571	-0.021066	-0.00026315
572	0.0087942	0.0025608
573	-0.0074645	0.0024372
574	-0.0018673	-0.00016757
575	0.0067567	0.0035695
576	-0.0047443	0.0028286
577	-0.00063514	3.2965e-06
578	-0.0013649	0.0022562
579	0.0056345	0.00069306
580	-0.0004438	-0.0020274
581	0.040031	0.00047291
582	0.037465	-0.0012478
583	0.0035034	0.0059373
584	-0.014882	0.00044388
585	-0.01485	0.0014426
586	-0.0050712	0.0033121
587	0.0005951	-0.00082808
588	-0.006036	-0.0030366
589	-0.0048856	-0.00017958
590	-0.0054578	-0.00070448
591	0.017606	-0.00075541
592	0.0016507	0.0053225
593	0.0090953	0.0012336
594	0.011062	0.00042296
595	0.023931	0.002713
596	0.0025003	0.00077841
597	-0.0016735	-0.00077732
598	0.00013926	-0.0014561
599	-0.0034187	0.0035633
600	0.0028037	-0.0011822
601	-0.0042863	0.0014841
602	-0.012801	0.00073768
603	0.0086556	0.0041065
604	-0.00090717	0.0078509
605	0.0038985	0.0053674
606	0.0036847	0.007401
607	0.011473	0.0048105
608	0.017821	0.0034438
609	0.01623	0.0051331
610	0.005442	0.0028126
611	0.0020892	0.00085778
612	0.0096176	0.0024502
613	-0.0015959	0.005051
614	0.0055754	0.0057409
615	-0.0092084	0.002048
616	-0.0051187	-0.0019265
617	-0.00058485	0.0013551
618	0.00028301	-0.0012293
619	-0.0078909	0.0010829
620	-0.0058362	-0.00012161
621	-0.016504	0.0011243
622	0.0094557	0.0020745
623	-0.0073204	0.00022824
624	0.0075009	0.0073377
625	-0.012896	-0.0029306
626	0.0066661	0.0078457
627	-0.0044561	-0.00098778
628	-0.015067	-0.0039066
629	-0.0073279	0.00026616
630	-0.0085295	-0.00018966
631	-0.0058183	-0.0022425
632	-0.0060026	-0.0010389
633	-0.0066174	-0.00090711
634	-0.0086327	-0.001891
635	-0.0042455	0.0013853
636	0.014907	-0.0019105
637	0.005363	0.0002558
638	-0.01979	-0.007683
639	0.0010922	-5.7423e-05
640	-0.0045941	-0.0071288
641	-0.012642	0.00050193
642	0.0040434	0.00011611
643	0.016229	0.0059111
644	-0.019546	0.001248
645	-0.021882	-0.0030746
646	-0.016894	-0.0014733
647	-0.01378	0.0013246
648	-0.0048314	0.0029808
649	-0.023119	-0.00028354
650	0.00069488	0.0027244
651	-0.0037764	-0.0039701
652	0.0068902	-0.0024492
653	0.0038457	0.0040525
654	-0.0032546	0.0028818
655	-0.013072	0.0030545
656	-0.014649	-0.0045223
657	-0.0094935	0.0018366
658	0.0018682	0.0055703
659	-0.0056431	0.0077361
660	-0.017274	0.0052928
661	-0.0029743	0.0046889
662	-0.031596	-0.0043854
663	0.004532	0.0032271
664	0.004709	0.0036704
665	-0.00087394	-0.00054243
666	0.003216	-0.0013641
667	0.01468	0.0059242
668	-0.013761	-0.00035058
669	-0.015741	0.00013806
670	0.006561	0.005592
671	0.0088076	0.0026754
672	0.012436	0.0013091
673	-0.0091477	-0.0013185
674	-0.00070312	0.0059531
675	-0.015054	0.0091074
676	0.0001519	-0.00039641
677	0.016461	0.0047704
678	0.016874	0.0031461
679	-0.013786	0.0029758
680	0.0054142	-0.00039682
681	0.0071771	-0.0021675
682	-0.016139	-0.0057218
683	-0.00051625	-0.00028676
684	0.016862	-0.0012013
685	-0.0032273	0.0099211
686	-0.0043747	0.0051795
687	-0.0096094	-0.0019217
688	-0.0037798	-0.00089329
689	-0.019674	-0.00048349
690	0.00028099	0.0035593
691	0.016032	0.0034485
692	-0.00024432	0.0057374
693	-0.0082986	-0.0029583
694	-0.0033462	0.0012757
695	-0.007032	-0.00052844
696	-0.0057592	-0.0011325
697	-0.014785	5.3261e-05
698	-0.01635	0.0014551
699	0.0011012	-0.00022025
700	-0.01684	-0.0035503
701	0.0031391	-0.002103
702	-0.0078567	6.7572e-05
703	-0.018317	-0.0031429
704	-0.015203	-0.00021205
705	-0.011691	-0.0021323
706	-0.018541	-0.0023954
707	-0.0080221	0.0080712
708	0.0050189	0.0043942
709	0.0017392	-0.0001808
710	-0.013312	0.0015751
711	0.0084936	0.0020579
712	-0.009572	0.0057991
713	-0.007687	2.0787e-05
714	-0.007803	-0.00052525
715	0.0034205	0.0017671
716	0.002148	-0.001853
717	-0.0046287	0.0034152
718	-0.0052892	2.0177e-05
719	0.019736	6.9255e-05
720	0.0029608	0.0049084
721	0.002918	0.0039021
722	-0.01829	-0.00023047
723	-0.018718	0.0011652
724	-0.030788	0.0047455
725	-0.0055527	-0.0014077
726	0.003913	0.0048344
727	0.00112	0.001478
728	0.0062776	-0.0021664
729	-0.021979	-0.0069522
730	-0.019348	-0.0023155
731	-0.01285	3.7546e-05
732	-0.0057438	-0.00031335
733	0.0045751	0.0022719
734	0.0096848	-0.0010067
735	-0.013644	0.0017762
736	-0.023649	-0.0017566
737	-0.018075	-0.00337
738	-0.0042294	0.0009045
739	0.0031825	0.0034399
740	-0.020678	-0.0044294
741	0.002354	-0.00071298
742	-0.016049	-0.0029677
743	-0.0070825	0.0068244
744	-0.0067827	0.0056081
745	-0.0038068	0.0029172
746	0.0050406	0.0037622
747	0.0042205	0.0013325
748	-0.0061604	-1.7255e-05
749	0.0052476	-0.00090132
750	-0.0036695	0.0062213
751	0.0056671	0.0070092
752	0.00060515	0.0062513
753	0.0011476	0.0033288
754	-0.0022067	0.0048994
755	-0.0011972	0.00038519
756	0.015105	0.0038062
757	-0.0087627	0.0025573
758	-0.0072753	0.0017747
759	-0.016212	0.0060844
760	-0.0073084	-0.0013259
761	-0.013669	0.00072931
762	0.015999	0.0046036
763	-0.0097871	-0.0001069
764	-0.0041104	0.00034982
765	-0.01164	-0.00067873
766	-0.0023127	0.0016877
767	-0.01299	-0.00077208
768	-0.015642	-0.0012741
769	0.0072423	0.0014447
770	-0.001141	0.0016497
771	0.0065077	0.0030577
772	-0.0031131	0.0042847
773	-0.0020096	-0.0023291
774	-0.0095687	-0.00061899
775	-0.023506	-0.0047216
776	0.0017643	0.0023236
777	0.0022819	0.0028472
778	-0.0048719	0.0052781
779	0.0016725	0.00092476
780	0.0080112	0.0061568
781	-0.0039456	0.00078025
782	-0.014919	0.001261
783	-0.015917	0.00054212
784	-0.0034787	0.0047854
785	-0.020689	-0.0031289
786	-0.0049584	0.0063291
787	-0.0066943	0.0017972
788	0.016307	0.0039805
789	0.013516	0.0015569
790	-0.013383	-0.001567
791	-0.0033943	-0.00040831
792	-0.014963	0.0015506
793	-0.015353	0.0021516
794	0.007812	0.0013583
795	0.0052685	-0.00073144
796	0.005992	0.0060641
797	0.0013852	0.0012776
798	0.0029223	0.004519
799	-0.014632	0.00098772
800	-0.0089857	-0.0014931

PCA explained variance — how many dimensions?

View explained variance values

Variance per displayed principal component
Component	Spectral variance
PC1	82.08%
PC2	7.54%
PC3	2.28%
PC4	1.97%
PC5	1.70%
PC6	0.98%
PC7	0.90%
PC8	0.52%
PC9	0.28%
PC10	0.19%

X-Y spectral correlation 7

Spectral association with LMA

Read Pearson correlation (−1 to +1) along the spectral axis (nm); the solid curve shows |r| and the dashed curve shows signed r.

Spectral association with WCf

Read Pearson correlation (−1 to +1) along the spectral axis (nm); the solid curve shows |r| and the dashed curve shows signed r.

Spectral association with WQ

Read Pearson correlation (−1 to +1) along the spectral axis (nm); the solid curve shows |r| and the dashed curve shows signed r.

Target	max \|r\|	axis @ max	mean \|r\|	\|r\| ≥ .5
LMA	0.856	2.33e+03	0.46	54.5%
WCf	0.738	1.81e+03	0.26	16.3%
WQ	0.855	1.88e+03	0.589	68.1%
An	0.288	1.44e+03	0.103	0.0%
PSI	0.251	1.66e+03	0.131	0.0%
WUEintr	0.32	1.65e+03	0.163	0.0%
WUEinst	0.338	1.61e+03	0.181	0.0%

Metric interpretation reference

Definitions, formulas and possible explanations for the reported values. A possible cause requires verification against acquisition records. Reading guide

Metric catalog 30

Look up each measured quantity, its calculation and the heuristic alert mapping before interpreting a high or low value.

Family	Metric	What it measures	Higher values suggest	Lower values suggest	Possible causes	Calculation / score
Data integrity	`Invalid/masked cell count (owner report)`	Null or masked cells in the evaluated population	To investigate	Complete according to the declared mask	Missing data or validity mask; not evidence of an acquisition cause	Native Preflight count of cells unavailable under x/null and valid_mask; denominator stays in native_metric.denominators.n_cells. Not legacy NaN ratio.
Data integrity	`NaN ratio`	Missing data	Corrupted spectrum	Full spectrum	Acquisition/export error	count(isnan(X)) / X.sizealert = min(1, nan_ratio / 0.05)
Data integrity	`Inf count`	Infinite values	Corruption	Normal	Invalid calculations	count(isinf(X))alert = min(1, inf_count / 1)
Data integrity	`Zero ratio`	Null columns or cells	Truncated spectrum	Normal	Export, saturation	count(X == 0) / count(finite X)alert = min(1, zero_ratio / 0.05)
Overall amplitude	`Mean reflectance`	Average level	Too light / background visible	Too dark	Background, geometry	mean(X finite)alert reuses baseline/shape drift because absolute reflectance ranges are technology-dependent
Overall amplitude	`Area under curve`	Overall intensity	Illumination difference	Normal	Probe distance	trapezoid(mean_spectrum, spectral_axis)alert reuses baseline/shape drift because area scale depends on axis and units
Overall amplitude	`Peak-to-peak (PTP)`	Dynamic range	Strong variability	Flat spectrum	Saturation	max(mean_spectrum) - min(mean_spectrum)alert increases when dynamic range is abnormally flat
Overall amplitude	`Variance`	Spectral variability	Normal or heterogeneous	Flat spectrum	Bad contact	var(X finite)alert increases when variance/dynamic range is abnormally flat
Noise	`Roughness sigma proxy`	High frequency noise	Noisy	Stable	Lamp, detector	median MAD(second derivative) * 1.4826 / sqrt(6)alert = noise_rms / signal_scale, saturated at 5%
Noise	`Roughness S/N proxy`	Signal quality	Good signal	Bad signal	Acquisition	mean(abs(X)) / noise_rmsalert decreases with SNR dB; >=40 dB is treated as low alert
Noise	`Bandwise roughness S/N proxy`	Localized noise	Reliable zone	Problem area	Detector	min(abs(mean_spectrum) / local second-derivative noise)alert decreases with worst-band SNR dB; >=35 dB is treated as low alert
Local artifacts	`Spike count`	Narrow peaks	Artifacts	Clean spectrum	Cosmic rays, splice	count robust outliers in second derivativealert follows spike_rate, saturated at 1%
Local artifacts	`Spike rate`	Peak density	Suspicious spectrum	Normal	Interpolation	spike_count / (n_samples * (n_features - 2))alert = min(1, spike_rate / 0.01)
Local artifacts	`Jump count`	Discontinuities	Detector connection	Continuous	Splice	count robust outliers in first derivativealert follows jump_rate, saturated at 1%
Local artifacts	`Jump rate`	Jump frequency	Spectral problem	Normal	Calibration	jump_count / (n_samples * (n_features - 1))alert = min(1, jump_rate / 0.01)
Local artifacts	`Clip fraction`	Saturation	Clipping	Normal	Detector saturated	fraction of finite cells equal to repeated min/max extremaalert = min(1, clip_fraction / 0.01)
Spectral form	`Baseline slope`	Overall slope	Drift	Stable	Illumination	linear slope of mean_spectrum over normalized axisalert = abs(slope / signal_scale), saturated at 0.5
Spectral form	`Curvature RMS`	Curvature	Unusual shape	Smooth	Background, splice	median RMS(second derivative per spectrum)alert = curvature_rms / signal_scale, saturated at 1%
Spectral form	`D1 RMS`	Local variability	Structured spectrum	Flat	Biology or artifact	median RMS(first derivative per spectrum)alert = d1_rms / signal_scale, saturated at 5%
Multivariate outliers	`PCA Q (SPE)`	Not explained by PCA	Atypical spectrum	Compliant	Artifact, mixture	p95(Q/SPE residual) / median(Q/SPE residual)alert = min(1, pca_q_ratio / 8)
Multivariate outliers	`Hotelling T²`	Extreme in PCA	Extreme but consistent	Central	Natural variability	p95(Hotelling T2) / median(Hotelling T2)alert = min(1, hotelling_t2_ratio / 8)
Multivariate outliers	`Mahalanobis H`	Distance to cloud	Global outlier	Normal population	Different domain	p95(sqrt(T2)) / median(sqrt(T2))alert = min(1, mahalanobis_h_ratio / 4)
Comparison to average spectrum	`RMS to mean spectrum`	Average distance	Different spectrum	Typical	Domain shift	p95 RMS distance to dataset mean spectrumalert = RMS_p95 / signal_scale, saturated at 25%
Comparison to average spectrum	`Spectral Angle Mapper (SAM)`	Difference in shape	Different shape	Similar	Background, geometry	p95 spectral angle to dataset mean spectrumalert = min(1, SAM_p95 / 0.35 rad)
Repeatability	`RMS intra-ID`	Reproducibility	Poor repeatability	Stable	Positioning	median RMS distance to repeated-sample centroidalert = RMS_intra_ID / signal_scale, saturated at 10%
Repeatability	`SAM intra-ID`	Shape variation	Unstable	Stable	Acquisition	median SAM to repeated-sample centroidalert = min(1, SAM_intra_ID / 0.15 rad)
Repeatability	`CV intra-ID`	Internal variability	Bad control	Stable	Operator	median within-ID band CValert = min(1, CV_intra_ID / 0.25)
Structure of the dataset	`PCA score density`	Clusters	Subpopulations	Homogeneous	Different batches	1 / median kNN distance in PCA score spacealert follows density_cv/profile structure complexity, not raw density alone
Structure of the dataset	`kNN distance ratio (not LOF)`	Local anomaly	Isolated spectrum	Normal population	Rare cases	p95(mean kNN distance / median population mean kNN distance) in PCA-score space; not LOFalert = min(1, max(0, kNN_distance_ratio_p95 - 1) / 2)
Structure of the dataset	`Isolation Forest score`	Global anomaly	Atypical spectrum	Normal	Various causes	p95 IsolationForest anomaly score on PCA scoresalert follows structure complexity; raw score is implementation-dependent

Technology-specific extensions

Find measurement-specific checks and acquisition context that can help interpret a source's spectral patterns.

Technology	Adaptations / metrics	Target anomalies	Practical guidance
UV-Vis 300-1000 nm	Baseline, overall slope, drift at edges 300-350 and 900-1000; metrics by zones	Stray light, bad white, saturation, weak signal at the ends	The edges are often unstable; also calculate edge/middle scores.
UV-Vis 300-1000 nm	Saturation / clipping near max absorbance or max reflectance	Signal clipped	Very important if strong absorption.
UV-Vis 300-1000 nm	Red-edge, peak position, band ratios for vegetation	Biological shift or optical artifact	Helps distinguish real change from acquisition problem.
UV-Vis 300-1000 nm	Smoothness / roughness index	High frequency noise	Often more informative than the SNR alone.
MIR / ATR-FTIR	ATR contact quality index: overall intensity, total area, depth of key bands	Poor crystal-sample contact	Crucial: many anomalies come from the ATR contact.
MIR / ATR-FTIR	CO2 / H2O atmospheric bands	Bad atmospheric correction	Frequent spurious peaks.
MIR / ATR-FTIR	Baseline curvature / rubber-band residual	Diffusion, contact, baseline drift	Very useful before PCA.
MIR / ATR-FTIR	Peak position shift	Bad spectral alignment/calibration	Important in FTIR because small shifts count.
MIR / ATR-FTIR	Band area ratios on known bands	Chemically incoherent spectrum	To be adapted by matrix: polysaccharides, proteins, lipids, etc.
HS-MS	Total Ion Current (TIC), Base Peak Intensity (BPI)	Weak injection, unstable ionization	MS equivalent of the global spectral level.
HS-MS	Number of peaks detected	Poor or too noisy spectrum	Too little = bad signal; too much = noise/contamination.
HS-MS	Mass accuracy / m/z drift	Mass calibration problem	Fundamental in HRMS.
HS-MS	Retention time drift if LC/GC-MS	Chromatographic drift	Monitor on standards/QC pools.
HS-MS	Blank contamination score	Contaminants / carry-over	Compare samples vs blanks.
HS-MS	Internal standard CV	Instrumental variability	Very robust if standards available.
HS-MS	Missingness by feature	Detection instability	Crucial for filtering variables.
With repetitions	Intra-sample RMS	Overall repeatability	Applicable to all technologies.
With repetitions	SAM / intra-sample correlation	Shape repeatability	Very useful for spectra.
With repetitions	Intra-sample CV per band / feature	Local repeatability	Detects unstable areas.
With repetitions	ICC or variance components	Share of sample vs. technical variance	Very useful if several repetitions per sample.
With repetitions	Distance to intra-ID centroid	Aberrant repetition	Allows you to flag the bad repetition rather than the entire sample.

Bug-hunting / supervised audits

Review suggested follow-up methods for suspected data issues and whether their results are available in this explorer.

Potential issue family	Suggested methods	What it detects	Explorer availability
Global spectral shift	Inter-dataset average spectrum correlation, DTW, cross-correlation, comparison of peak positions	Wavelength shift, misalignment, different interpolation	Partially calculated: cross-correlation lag and dispersion of peak positions vs. average spectrum.
Baseline / offset / gain	Regression each spectrum vs average spectrum: x = a + b ref + residual; followed by a, b, residual RMS	Additive offset, multiplicative effect, baseline drift	Calculated in reference.affine_*.
Row mixing / X-M-Y mismatch	Index check, row hash, ID duplication, intra-ID spectral distance, inconsistent labels	Mixed up rows, misaligned metadata, Y assigned to wrong spectrum	Partially covered by intra-ID repeatability; checks index/hash to add to the canonical pipeline.
Information leak / poorly split repetitions	GroupKFold by sample_id vs StratifiedKFold random; auditing partitions by sample_id	Artificially good performance due to repetitions	Requires splits and benchmark model; not calculated by the descriptive card.
Label bugs	Close samples in X but different Y, confident learning, systematic FP/FN errors	Inverted Y, input errors, ambiguous classes	Requires Y and/or model; recommended for supervised explorer.
Hidden subdomains	PCA/UMAP/t-SNE + unsupervised clustering + association with dataset/Y/date/operator	Batches, campaigns, probes, backgrounds not provided	Partially calculated by PCA and distance ratio kNN (not LOF); UMAP/t-SNE outside the static card.
Unknown localized artifacts	Wavelength x dataset map: mean difference, variance difference, KS per wavelength	Unanticipated anomalous spectral regions	To be calculated at the bank level when several datasets share a spectral axis.
Instrumental breaks	Discontinuities in derivatives, changepoint detection	Splice, detector connection, local jump not provided	Calculated by jump/spike rates; more advanced changepoint to add.
Spectral mixing/contamination	NMF / unmixing / reconstruction by convex hull	External component: background, plastic, soil	Not calculated automatically; requires component assumptions or large library.
Unstable but predictive features	Model importance vs. QC instability by variable	Model that learns an artifact rather than a biological signal	Requires supervised model; recommended for benchmark reporting.

Curated scientific variables

Scientific roles derived from the source contract. M denotes experimental metadata; Y denotes informative response variables. Reading guide

14 experimental metadata fields (M) · 7 informative targets (Y).

Variable	Role	Distinct values	Complete	Range / balance
instrument_name	M	3	100%	IR 1.138
experiment	M	2	100%	IR 2.097
phenotyping_type	M	2	100%	IR 18.94
variety_code	M	246	67.68%	IR 10.75
water_treatment	M	3	67.68%	IR 1.025
date_leaf_measurement	M	23	61.15%	IR 30.67
nb_min_since_8am	M	1,051	61.15%	IR 5
nb_min_since_last_irrigation	M	1,259	60.85%	IR 3
temp_mean_arch	M	72	61.15%	IR 92
hr_mean_arch	M	72	61.15%	IR 92
ppfd_mean_arch	M	72	61.15%	IR 92
vpd_mean_arch	M	72	61.15%	IR 92
duration_from_entry_to_measure	M	37	61.15%	IR 156
plant_density_mean	M	518	61.02%	IR 69
lma	Y	1,249	99.28%	1.429 – 9.251
wcf	Y	1,436	62.67%	0.6061 – 0.8584
wq	Y	1,048	62.67%	41.1 – 189.5
an	Y	120	5.015%	0.05647 – 18.62
psi	Y	113	4.9%	-1.966 – -0.26
wueintr	Y	117	4.885%	14.13 – 187.6
wueinst	Y	117	4.885%	0.8329 – 12.39

Identifiers and sampling groups remain available separately. Statistics count exported observations.

Variables

Declared prediction targets and recorded metadata, with distributions and missing-value counts. Identifiers are not automatically prediction targets. Reading guide

Targets 7

LMA

target · numeric

Read the recorded value distribution and summary statistics; no measurement unit is recorded.

n / missing ? Recorded observation count and number of missing values for this variable.	6,919 / 50
Mean ± SD ? Arithmetic mean and standard deviation in the variable's stated unit.	4.591 ± 0.959
Median ? 50th percentile: half the non-missing values lie at or below it.	4.576
Range ? Smallest and largest recorded non-missing values.	1.429 – 9.251
CV ? Coefficient of variation: standard deviation divided by absolute mean; unstable when the mean is near zero.	0.209
Skew / kurtosis ? Distribution asymmetry and excess kurtosis. A Gaussian distribution has skewness 0 and excess kurtosis 0.	0.21 / 0.049
Normal? ? Descriptive normality-test result. A non-normal target is not automatically a defective variable.	no

WCf

target · numeric

Read the recorded value distribution and summary statistics; no measurement unit is recorded.

n / missing ? Recorded observation count and number of missing values for this variable.	6,919 / 2,583
Mean ± SD ? Arithmetic mean and standard deviation in the variable's stated unit.	0.7293 ± 0.0317
Median ? 50th percentile: half the non-missing values lie at or below it.	0.7284
Range ? Smallest and largest recorded non-missing values.	0.6061 – 0.8584
CV ? Coefficient of variation: standard deviation divided by absolute mean; unstable when the mean is near zero.	0.0435
Skew / kurtosis ? Distribution asymmetry and excess kurtosis. A Gaussian distribution has skewness 0 and excess kurtosis 0.	0.15 / 0.5
Normal? ? Descriptive normality-test result. A non-normal target is not automatically a defective variable.	no

WQ

target · numeric

Read the recorded value distribution and summary statistics; no measurement unit is recorded.

n / missing ? Recorded observation count and number of missing values for this variable.	6,919 / 2,583
Mean ± SD ? Arithmetic mean and standard deviation in the variable's stated unit.	86.4 ± 17
Median ? 50th percentile: half the non-missing values lie at or below it.	85.2
Range ? Smallest and largest recorded non-missing values.	41.1 – 189.5
CV ? Coefficient of variation: standard deviation divided by absolute mean; unstable when the mean is near zero.	0.197
Skew / kurtosis ? Distribution asymmetry and excess kurtosis. A Gaussian distribution has skewness 0 and excess kurtosis 0.	0.94 / 2.6
Normal? ? Descriptive normality-test result. A non-normal target is not automatically a defective variable.	no

An

target · numeric

Read the recorded value distribution and summary statistics; no measurement unit is recorded.

n / missing ? Recorded observation count and number of missing values for this variable.	6,919 / 6,572
Mean ± SD ? Arithmetic mean and standard deviation in the variable's stated unit.	6.563 ± 3.8
Median ? 50th percentile: half the non-missing values lie at or below it.	6.375
Range ? Smallest and largest recorded non-missing values.	0.05647 – 18.62
CV ? Coefficient of variation: standard deviation divided by absolute mean; unstable when the mean is near zero.	0.579
Skew / kurtosis ? Distribution asymmetry and excess kurtosis. A Gaussian distribution has skewness 0 and excess kurtosis 0.	0.58 / -0.024
Normal? ? Descriptive normality-test result. A non-normal target is not automatically a defective variable.	no

PSI

target · numeric

Read the recorded value distribution and summary statistics; no measurement unit is recorded.

n / missing ? Recorded observation count and number of missing values for this variable.	6,919 / 6,580
Mean ± SD ? Arithmetic mean and standard deviation in the variable's stated unit.	-0.9718 ± 0.391
Median ? 50th percentile: half the non-missing values lie at or below it.	-0.916
Range ? Smallest and largest recorded non-missing values.	-1.966 – -0.26
CV ? Coefficient of variation: standard deviation divided by absolute mean; unstable when the mean is near zero.	0.403
Skew / kurtosis ? Distribution asymmetry and excess kurtosis. A Gaussian distribution has skewness 0 and excess kurtosis 0.	-0.44 / -0.64
Normal? ? Descriptive normality-test result. A non-normal target is not automatically a defective variable.	no

WUEintr

target · numeric

Read the recorded value distribution and summary statistics; no measurement unit is recorded.

n / missing ? Recorded observation count and number of missing values for this variable.	6,919 / 6,581
Mean ± SD ? Arithmetic mean and standard deviation in the variable's stated unit.	90.26 ± 36.5
Median ? 50th percentile: half the non-missing values lie at or below it.	88.43
Range ? Smallest and largest recorded non-missing values.	14.13 – 187.6
CV ? Coefficient of variation: standard deviation divided by absolute mean; unstable when the mean is near zero.	0.404
Skew / kurtosis ? Distribution asymmetry and excess kurtosis. A Gaussian distribution has skewness 0 and excess kurtosis 0.	0.47 / -0.092
Normal? ? Descriptive normality-test result. A non-normal target is not automatically a defective variable.	no

WUEinst

target · numeric

Read the recorded value distribution and summary statistics; no measurement unit is recorded.

n / missing ? Recorded observation count and number of missing values for this variable.	6,919 / 6,581
Mean ± SD ? Arithmetic mean and standard deviation in the variable's stated unit.	4.319 ± 1.98
Median ? 50th percentile: half the non-missing values lie at or below it.	3.998
Range ? Smallest and largest recorded non-missing values.	0.8329 – 12.39
CV ? Coefficient of variation: standard deviation divided by absolute mean; unstable when the mean is near zero.	0.459
Skew / kurtosis ? Distribution asymmetry and excess kurtosis. A Gaussian distribution has skewness 0 and excess kurtosis 0.	1.4 / 2.9
Normal? ? Descriptive normality-test result. A non-normal target is not automatically a defective variable.	no

Metadata 17

axis_min

metadata · numeric

Read the recorded value distribution and summary statistics; no measurement unit is recorded.

n / missing ? Recorded observation count and number of missing values for this variable.	6,919 / 0
Mean ± SD ? Arithmetic mean and standard deviation in the variable's stated unit.	867.4 ± 384
Median ? 50th percentile: half the non-missing values lie at or below it.	908.1
Range ? Smallest and largest recorded non-missing values.	400 – 1350
CV ? Coefficient of variation: standard deviation divided by absolute mean; unstable when the mean is near zero.	0.443
Skew / kurtosis ? Distribution asymmetry and excess kurtosis. A Gaussian distribution has skewness 0 and excess kurtosis 0.	-0.016 / -1.5
Normal? ? Descriptive normality-test result. A non-normal target is not automatically a defective variable.	no

axis_max

metadata · numeric

Read the recorded value distribution and summary statistics; no measurement unit is recorded.

n / missing ? Recorded observation count and number of missing values for this variable.	6,919 / 0
Mean ± SD ? Arithmetic mean and standard deviation in the variable's stated unit.	2229 ± 404
Median ? 50th percentile: half the non-missing values lie at or below it.	2,500
Range ? Smallest and largest recorded non-missing values.	1676 – 2550
CV ? Coefficient of variation: standard deviation divided by absolute mean; unstable when the mean is near zero.	0.181
Skew / kurtosis ? Distribution asymmetry and excess kurtosis. A Gaussian distribution has skewness 0 and excess kurtosis 0.	-0.63 / -1.6
Normal? ? Descriptive normality-test result. A non-normal target is not automatically a defective variable.	no

n_points_original

metadata · numeric

Read the recorded value distribution and summary statistics; no measurement unit is recorded.

n / missing ? Recorded observation count and number of missing values for this variable.	6,919 / 0
Mean ± SD ? Arithmetic mean and standard deviation in the variable's stated unit.	849.4 ± 912
Median ? 50th percentile: half the non-missing values lie at or below it.	257
Range ? Smallest and largest recorded non-missing values.	125 – 2,101
CV ? Coefficient of variation: standard deviation divided by absolute mean; unstable when the mean is near zero.	1.07
Skew / kurtosis ? Distribution asymmetry and excess kurtosis. A Gaussian distribution has skewness 0 and excess kurtosis 0.	0.64 / -1.6
Normal? ? Descriptive normality-test result. A non-normal target is not automatically a defective variable.	no