CODEWITHSUNDEEP
pythonpandas
bewolf
bewolf

Reputation: 165

Remove raws when one one sequence is in a cluster in df

I have a data frame such as:

Cluster sequence_name
1   specie1
1   specie2
1   specie3
1   sequence1
1   sequence2
2   specie8
3   specie2
4   sequence1
4   sequence3
4   specie56
...

I would like to remove all the cluster that contain only one sequence, here in the exemple I should get: 

Cluster sequence_name
1   specie1
1   specie2
1   specie3
1   sequence1
1   sequence2
4   sequence1
4   sequence3
4   specie56
...

Thank you for your help .

Upvotes: 1

Views: 28

Answers (2)

Vaishali
Vaishali

Reputation: 38415

Groupby.filter works well here

df = df.groupby('Cluster').filter(lambda x: x.sequence_name.nunique() > 1)

    Cluster sequence_name
0   1       specie1
1   1       specie2
2   1       specie3
3   1       sequence1
4   1       sequence2
7   4       sequence1
8   4       sequence3
9   4       specie56

Upvotes: 1

It_is_Chris
It_is_Chris

Reputation: 14103

Boolean indexing with groupby and transform:

df[df.groupby('Cluster')['sequence_name'].transform('size') > 1]

   Cluster sequence_name
0        1       specie1
1        1       specie2
2        1       specie3
3        1     sequence1
4        1     sequence2
7        4     sequence1
8        4     sequence3
9        4      specie56

Upvotes: 1

Related Questions