krank.fetch_liwc

krank.fetch_liwc(dataset, version=None, load=True, target_dic=None, **kwargs)

Fetch a LIWC scores dataset as a pandas.DataFrame.

Calling this function will always retrieve/download the file from its remote source if it has not already been retrieved (via pooch.retrieve()).

Refer to the online krank.liwc documentation for more details about the module and available LIWC scores datasets.

Note

Custom loading

To apply custom loading, either (1) set load=False to return the local filepath and load the file in subsequent steps, or (2) pass a callable function to load and it will be used to load the file instead of the krank default loader.

See also

This is a top-level wrapper around dataset-specific functions in the krank.lexicons module. For example, calling krank.fetch_liwc("barrett2020") is identical to calling krank.tables.fetch_barrett2020().

Parameters

datasetstr

Name of dataset (i.e., lexicon). Refer to the lexicons documentation for a list of all available datasets.

versionstr or None

Name of version. If None (default), fetches the latest version. Refer to the lexicons documentation for a list of all available versions within each dataset.

loadbool or callable

If False (default), fetch the file and return the local filepath. If True, fetch the file and load it as a pandas.DataFrame. If a callable, fetch the file an load it with the custom callable.

target_dicstr or None

If None (default), LIWC categories remain in their original format. If not None, LIWC categories are converted from the source dictionary (whatever was used in the original study) to the specified target_dic. If target_dic is the same as the source dictionary, no action is taken. If not None, must be one of LIWC1999, LIWC2001, LIWC2007, LIWC2015, or LIWC-22.

Warning

This feature is experimental.

Categories change between versions of LIWC, and so these are far from comparable. Some categories will not be present in both source and target dictionaries, and thus removed from the returned dataframe. Other categories might have close-but-imperfect fits or might have changed drastically between LIWC versions. In this case, the conversions are still provided, so be careful with interpretations. A useful resource if wanting to use scores pooled across LIWC versions would be the LIWC manuals, where comparisons between LIWC versions are described in detail and analyses comparing output across LIWC versions are presented.

**kwargsdict, optional

Additional keyword arguments are passed to pooch.retrieve().

Returns

str or DataFrame

Path of retrieved file if load is False, or pandas.DataFrame if load is True.