gaussian.diagnostics

Verification datasets: sample once, reuse across metrics and plots.

Functions

verification_settings(context)
normalization_settings(context)
model_evaluation_settings(context)
test_predictions(context, dependencies, writer)
verification_observations(context, dependencies, writer)
verification_predictions(context, dependencies, writer)
sliced_wasserstein(source, target, weights, directions)	Mean exact weighted 1D W1 over supplied unit directions (not a histogram approximation).
reweighting_distances(context, dependencies, writer)
normalization_observations(context, dependencies, writer)
normalization_predictions(context, dependencies, writer)
showcase_observations(context, dependencies, writer)
showcase_predictions(context, dependencies, writer)

Module Contents

gaussian.diagnostics.verification_settings(context)

gaussian.diagnostics.normalization_settings(context)

gaussian.diagnostics.model_evaluation_settings(context)

gaussian.diagnostics.test_predictions(context, dependencies, writer)

gaussian.diagnostics.verification_observations(context, dependencies, writer)

gaussian.diagnostics.verification_predictions(context, dependencies, writer)

gaussian.diagnostics.sliced_wasserstein(source, target, weights, directions)

Mean exact weighted 1D W1 over supplied unit directions (not a histogram approximation).

gaussian.diagnostics.reweighting_distances(context, dependencies, writer)

gaussian.diagnostics.normalization_observations(context, dependencies, writer)

gaussian.diagnostics.normalization_predictions(context, dependencies, writer)

gaussian.diagnostics.showcase_observations(context, dependencies, writer)

gaussian.diagnostics.showcase_predictions(context, dependencies, writer)