CODEWITHSUNDEEP
pythonpandas
Neil
Neil

Reputation: 8247

how to find one value is associated with multiple values in pandas

I have following dataframe in pandas

  code     tank      product
  1234     1         MS
  1234     2         HS
  1234     1         HS
  1234     1         HS
  1235     1         MS
  1235     1         HS
  1235     1         MS
  1245     1         MS
  1245     2         HS

I want to find how many tanks have multiple products associated with them, in above dataframe e.g for code 1234 tank 1 has Ms and HS as well There are 2 cases in above dataframe

My Desired Dataframe would be

  code     tank     flag
  1234     1        yes
  1234     2        no
  1235     1        yes
  1245     1        no
  1245     2        no

How can I do it in pandas?

Upvotes: 2

Views: 753

Answers (1)

jezrael
jezrael

Reputation: 862611

Use SeriesGroupBy.nunique for count unique values per groups:

df = df.groupby(['code','tank'])['product'].nunique().reset_index()
print (df)
   code  tank  product
0  1234     1        2
1  1234     2        1
2  1235     1        2
3  1245     1        1
4  1245     2        1

And then extract column with pop and set values by numpy.where

df['flag'] = np.where(df.pop('product') == 1, 'no', 'yes')
print (df)
   code  tank flag
0  1234     1  yes
1  1234     2   no
2  1235     1  yes
3  1245     1   no
4  1245     2   no

Upvotes: 4

Related Questions