Count occurrences of certain string in entire pandas dataframe

Question

I have following dataframe in pandas

 C1      C2       C3
 10      a        b
 10      a        b
 ?       c        c
 ?       ?        b
 10      a        b
 10      ?        ?

I want to count the occurrences of ? in all the columns

My desired output is column wise sum of occurrences

Possible duplicate of [How to get value counts for multiple columns at once in Pandas DataFrame?](https://stackoverflow.com/questions/32589829/how-to-get-value-counts-for-multiple-columns-at-once-in-pandas-dataframe) — busybear, Feb 15 '19 at 17:23
Might help to make it easy for people and provide the full dataframe definition. — Tammo Heeren, Feb 15 '19 at 17:25
@busybear I do not want count of all unique occurrences in the dataframe. I want specific `?` occurrences in all columns. — Neil, Feb 15 '19 at 17:25

anky · Accepted Answer · 2019-02-15T17:32:13.880

4

Use:

m=df.eq('?').sum()
pd.DataFrame([m.values],columns=m.index)

   C1  C2  C3
0   2   2   1

Or better :

df.eq('?').sum().to_frame().T #thanks @user3483203

C1  C2  C3
0   2   2   1

edited Feb 15 '19 at 17:32

answered Feb 15 '19 at 17:26

anky

1

`df.eq('?').sum().to_frame().T` should be a bit faster. – user3483203 Feb 15 '19 at 17:30

1 Answers1