precomputed mean,std statistics using scipy.stats.ttest_ind_from_stats

[pas-calc]

If the test was performed on mean,std (if we don't have the raw samples, or it's a weighted mean,std) then we can use
scipy.stats.ttest_ind_from_stats

https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.ttest_ind_from_stats.html

But how to implement this in statannotations?
We don't have a dataframe with the samples data.

Related (seaborn)
mwaskom/seaborn#3345

[phaleanopsis]

It is always possible to provide custom p-values. In cases such as this I think it is generally easier to calculate p-values using statistics packages and then using statannotations only for the annotation.

annotator.configure(**statconfig_kwarg)
annotator.set_pvalues([pvalue])
annotator.annotate()