API Reference

Diagnostics API

overview(data) Give a brief data overview.
warnings(data[, corr_thresh, …]) Warn about common dataset problems.
column_report(data) Give column summaries according to pandas dtype.
profile(data[, corr_thresh, missing_thresh, …]) Profile dataset.

Selection API

RandomSelect([names, n_feats]) Randomly choose a feature set.
Dendrogram([X, pairing_func, max_threshes]) Pair features by an arbitrary function.
Dendrogram.fit(X[, pairing_func, max_threshes]) Build graphs for a given pairing function.
Dendrogram.transform(X[, n_feats]) Return a dataframe of a particular size.
Dendrogram.set_params(**params) Method to functionally assign parameters.
Dendrogram.features_at_step(step) Find the representatives at a certain step for a given graph.
Dendrogram.find_set_of_size(size) Finds a column set of a certain size in the Dendrogram.
Dendrogram.score_at_point(X, y, model, …) A helper method for scoring a Dendrogram at a step.
Dendrogram.shuffle_score_at_point(X, y, …) A helper method for scoring a Dendrogram at a step.

Learning API

create_model(X, y[, model, metric, …]) Make a model.
inplace_encoder(X) Replace all columns with pd.dtype == ‘O’ with integers.
feature_importances(X, model[, n_feats]) Print a list of important features.
create_holdout(X, y[, split_size]) A wrapper around train_test_split.

Plotting API

show(plot[, png, static, hover, width, …]) Format and show a bokeh plot.
show_template() Prints a template for show.
piechart(col[, sort, mergepast, drop_n, figargs]) Creates a piechart.
histogram(col[, y, n_bins, col_max, …]) Creates a histogram.
timeseries(col_1, col_2[, col_max, col_min, …]) Creates a time based aggregations of a numeric variable.
scatter(col_1, col_2[, cat, label, …]) Creates a scatter plot of two variables.
dendrogram(D[, figargs]) Creates a dendrogram plot.
feature_importances(X, model[, n_feats, figargs]) Plot feature importances.
roc_auc(X, y, model[, pos_label, prob_col, …]) Plots the reveiver operating characteristic curve.
f1(X, y, model[, n_precs, n_splits, figargs]) Plots the precision, recall and f1 at various thresholds.