Python: How can I filter DataFrame by type of data in pandas?

Question

I have a DataFrame like this;

df = pd.DataFrame([
    ['A1', ['Long Sleeves', 'Jacket'], 85],
    ['B1', 'Shoes', 55],
    ['A2', 'Skirt', 40]
], columns=['PRDS_ID', 'CAT', 'PRICE'])

I want to select rows that the type is list from column 'CAT'.
Therefore, I tried following codes;

df[df.CAT.astype(list) == True]

df['CAT'].apply(lambda x: len(x) > 1)

but neither of them worked. The first row should be returned in this case. How can I fix it?

Answer 1

Is that what you mean?

df['CAT'].apply(lambda x: x if isinstance(x, list) else None)

0    [Long Sleeves, Jacket]
1                      None
2                      None
Name: CAT, dtype: object

Answer 2

You can also use apply like

df[df.CAT.apply(type) == list]
  PRDS_ID                     CAT  PRICE
0      A1  [Long Sleeves, Jacket]     85

• df.CAT.apply(type) Will give you Series of object types of each rows then
• By comparing df.CAT.apply(type) == list we can get booleans to do indexing

Answer 3

PANDAS cells do not have an istype method; that's for Python base objects and others that implement the operator. Instead, you have to wrap your check inside a lambda function that you apply to the contents of each cell: you refer to the Python object inside the cell, and test that .

>>> df[df.iloc[:,1].map(lambda x: type(x) == list)]
  PRDS_ID                     CAT  PRICE
0      A1  [Long Sleeves, Jacket]     85
>>> df[df.iloc[:,1].map(lambda x: type(x) == str)]
  PRDS_ID    CAT  PRICE
1      B1  Shoes     55
2      A2  Skirt     40