pandas unique values multiple columns different dtypes

Question

Similar to pandas unique values multiple columns I want to count the number of unique values per column. However, as the dtypes differ I get the following error:

The data frame looks like A small[['TARGET', 'title']].apply(pd.Series.describe) gives me the result, but only for the category types and I am unsure how to filter the index for only the last row with the unique values per column

Is `small[['TARGET','title']].apply(lambda x: np.unique(x).size)` what you're after? — Abdou, Oct 08 '16 at 14:09
Indeed. Thanks. Please feel free to write this as the answer — Georg Heiler, Oct 08 '16 at 14:30

score 1 · Accepted Answer · answered Oct 08 '16 at 14:32

1

Use apply and np.unique to grab the unique values in each column and take its size:

small[['TARGET','title']].apply(lambda x: np.unique(x).size)

Thanks!

answered Oct 08 '16 at 14:32

Abdou

12,931
4
39
42

pandas unique values multiple columns different dtypes

1 Answers1