Since polars 0.20 the output of polars.Series.value_counts names the counts column count whereas it previously was counts. This renames to counts afterwards to be compatible with the other functions that depend on that time.
Also fixes some losses that used df.group_by(other_df[col]) since that doesn't seem to work anymore, even though the df and the series have the same number of rows.
Since polars 0.20 the output of
polars.Series.value_counts
names the counts columncount
whereas it previously wascounts
. This renames tocounts
afterwards to be compatible with the other functions that depend on that time.Also fixes some losses that used
df.group_by(other_df[col])
since that doesn't seem to work anymore, even though the df and the series have the same number of rows.