Pandas: boolean indexing with 'item in list' syntax

Question

Say I have a DataFrame with a column called col1. If I want to get all rows where col1 == ‘a’, I can do that with:

df[df.col1 == ‘a’]

If I want rows where col1 is ‘a’ or ‘b’, I can do:

df[(df.col1 == ‘a’) | (df.col1 == ‘b’)]

But I’d really like to do the is something (syntactically illegal) like this:

df[df.col1 in [‘a’, ‘b’, ‘c’]]

Is there a proper pandas way to do that?

Here’s what I’m using instead:

sort_func = lambda x: x in [‘a’, ‘b’, ‘c’]
mask = df[‘col1’].apply(sort_func)
df[mask]

But… is there a better way to do this? This is bothering me.

Answer 1

Use isin() for filtering

In [212]: df = pd.DataFrame([['a', 1], ['b', 2], ['c', 3], ['d', 4]],
                            columns=['col1', 'col2'])


In [213]: df['col1'].isin(['a', 'b', 'c'])
Out[213]:
0     True
1     True
2     True
3    False
Name: col1, dtype: bool

In [214]: df.loc[df['col1'].isin(['a', 'b', 'c']), :]
Out[214]:
  col1  col2
0    a     1
1    b     2
2    c     3