ModelDiagnostic

class protopipe.benchmarks.ModelDiagnostic(model, feature_name_list, target_name)

Bases: object

Base class for model diagnostics.

Parameters

model: `~sklearn.base.BaseEstimator`

Best model

feature_name_list: list

List of the features used to buil the model

target_name: str

Name of the target (e.g. score, gamaness, energy, etc.)

Methods Summary

add_image_model_output()
plot_feature_importance(ax, **kwargs)	Plot importance of features
plot_features(suptitle, data_list[, nbin, ...])	Plot model features for different data set (e.g.

Methods Documentation

add_image_model_output()

plot_feature_importance(ax, **kwargs)

Plot importance of features

Parameters

ax: `~matplotlib.axes.Axes`

Axis

plot_features(suptitle, data_list, nbin=30, hist_kwargs_list={}, error_kw_list={}, ncols=2)

Plot model features for different data set (e.g. training and test samples).

Parameters

data_list: list

List of data

nbin: int

Number of bin

hist_kwargs_list: dict

Dictionary with histogram options

error_kw_list: dict

Dictionary with error bar options

ncols: int

Number of columns

previous

ClassifierDiagnostic

next

OnlineBinnedStats