smx.plotting.threshold#

plot_threshold_spectrum: visualise a multivariate threshold overlaid on the original spectral zone, coloured by class.

Requires plotly. The dependency is optional — import errors produce a clear, actionable message rather than a hard package-load failure.

Functions#

plot_threshold_spectrum(→ Union[None, pandas.Series])

Reconstruct a threshold to spectrum space and save an HTML plot.

Module Contents#

smx.plotting.threshold.plot_threshold_spectrum(lrc_natural_df: pandas.DataFrame, row_index: int, spectral_zones_original: Dict[str, pandas.DataFrame], pca_info_dict_original: Dict, y_labels: pandas.Series, output_path: str | pathlib.Path | None, class_colors: Dict[str, str] | None = None, theme: smx.plotting.theme.SMXTheme | None = None, width: int | None = 900, height: int | None = 450, return_df: bool = False) → None | pandas.Series[source]#

Reconstruct a threshold to spectrum space and save an HTML plot.

The plot overlays the reconstructed multivariate threshold (in red) on top of the individual sample spectra for the chosen spectral zone, coloured by class label. The figure is always displayed; set return_df=True to return the threshold spectrum Series.

Parameters#

lrc_natural_dfpd.DataFrame: LRC DataFrame with natural-scale thresholds. Must contain columns 'Zone', 'Threshold_Natural', and 'Node_Natural'.
row_indexint: Row of lrc_natural_df to visualise.
spectral_zones_originaldict[str, pd.DataFrame]: Spectral zones extracted from the unpreprocessed calibration data.
pca_info_dict_originaldict: PCA info from smx.zones.aggregation.ZoneAggregator fitted on the natural (unpreprocessed) data.
y_labelspd.Series: Class labels aligned with the calibration data rows.
output_pathstr or Path, optional: Destination path for the output .html file. If None, no file is written.
class_colorsdict, optional: Mapping of class label → colour string. Explicit values override the theme. Defaults to the theme’s class_colors.
themeSMXTheme, optional: Visual theme. Defaults to smx.plotting.theme.DEFAULT_THEME.
return_dfbool, default False: If True, return the threshold spectrum Series.

Raises#

ImportError: If plotly is not installed.