ReliabilityModelPlot
- class ReliabilityModelPlot[source]
Bases:
objectVisualize
ReliabilityModeloutputs — calibration and the two trust axes.Turns the tables from
ReliabilityModelinto a per-sample view — each prediction’s score with its uncertainty interval, colored by trust status (ranking()) — and three batch diagnostics: a calibration curve (reliability_diagram(), fromReliabilityModel.eval()), the out-of-distribution score distribution (ood_hist()), and a score-vs-OOD map colored by thereliableflag (trust_map(); the last three fromReliabilityModel.predict()).Warning
Experimental. This class is part of the new v1.1.0 prediction layer and is under active development; its API (signatures, defaults, return objects) may change between minor releases without the usual deprecation cycle. Pin a version if you depend on the current behaviour.
Added in version 1.1.0.
See also
ReliabilityModelfor the measures these methods visualize.
Methods
ood_hist(df_rel, *[, figsize, bins, color, ...])Histogram of the out-of-distribution score, with the in-domain boundary at
1.0.ranking(df_rel, *[, names, figsize, top_n, ...])Per-sample view — each prediction's score with its uncertainty, colored by trust status.
reliability_diagram(df_eval, *[, figsize, ...])Calibration curve — mean predicted score vs.
trust_map(df_rel, *[, figsize, title, ax])Score vs.