What is the default scores used in the html created by scattertext.produce_scattertext_explorer function?

[AnnaDai1001]

Hello, I have been trying to figure out what are the default "scores" used when creating HTML file by scattertext.produce_scattertext_explorer function and read through the source code for hours but cant figure it out. Could anyone help me with this? Really appreciate it. I have created the html with the following piece of code:

        html = st.produce_scattertext_explorer(corpus,
                                                category='Positive', 
                                                category_name='Positive', 
                                                not_category_name='Negative',
                                                width_in_pixels=1000)

This means that I didn't specify the parameter scores so the default will be scores=None based on the source code. Plus, I didn't specify "term_scorer" either so the default will be term_scorer=None. From the source code of function produce_scattertext_explorer we have below. So I think scores will still be None. But in the HTML file, the terms are ranked by some scores. I am wondering what are these scores then? I have tried to calculate different metrics in each category, e.g. the frequency, f scaled score, pos precision etc. but none of them matched the HTML file.

    if term_scorer:
        scores = get_term_scorer_scores(category, corpus, neutral_categories, not_categories, show_neutral, term_ranker,
                                        term_scorer, use_non_text_features)

Really appreciate your help!

[JasonKessler]

The RankDifference class is used. This part of the package is admittedly quite shambolic.

See https://nbviewer.jupyter.org/github/JasonKessler/PuPPyTalk/blob/master/notebooks/Class-Association-Scores.ipynb for an explanation of this metric.