Utils

General utilities for xlogit library.

xlogit.utils.wide_to_long(dataframe, id_col, alt_list, alt_name, varying=None, sep='_', alt_is_prefix=False, empty_val=nan)

Reshapes pandas DataFrame from wide to long format.

Parameters

• dataframe (pandas DataFrame) – The wide-format DataFrame.

• id_col (str) – Column that uniquely identifies each sample.

• alt_list (list-like) – List of choice alternatives.

• alt_name (str) – Name of the alternatives column in returned dataset.

• varying (list-like) – List of column names that vary across alternatives.

• sep (str, default='_') – Separator of column names that vary across alternatives.

• avail (array-like, shape (n_samples,), default=None) – Availability of alternatives for the choice situations. One when available or zero otherwise.

• alt_is_prefix (bool) – True if alternative is prefix of the variable name or False if it is suffix.

• empty_val (int, float or str, default=np.nan) – Value to fill when alternative not available for a certain variable.

Returns

Return type

DataFrame in long format.

xlogit.utils.lrtest(general_model, restricted_model)

Conducts likelihood-ratio test.

Parameters

• general_model (xlogit Model) – Fitted model that contains all parameters (unrestricted)

• restricted_model (xlogit Model) – Fitted model with less parameters than general_model.

Returns

lrtest_result – p-value result, chisq statistic, and degrees of freedom used in test

Return type

dict