问题
What is the fastest way to drop columns[3] and columns[9:15]?
(I'm only able to remove the columns in 2 steps using the df.drop method)
1 2 3 4 5 6 .. n
A x x x x x x .. x
B x x x x x x .. x
C x x x x x x .. x
回答1:
You can, in fact, use pd.DataFrame.drop in one step. You can use np.r_ to combine multiple indices and ranges. Here's a demo:
df = pd.DataFrame(np.random.random((3, 20)))
print(df.columns) # RangeIndex(start=0, stop=20, step=1)
res = df.drop(np.r_[3, 9:15], 1)
print(res.columns)
# Int64Index([0, 1, 2, 4, 5, 6, 7, 8, 15, 16, 17, 18, 19], dtype='int64')
回答2:
Use as below:
>>> df
A B C D
0 0 1 2 3
1 4 5 6 7
2 8 9 10 11
>>> df.drop(['B', 'C'], axis=1)
A D
0 0 3
1 4 7
2 8 11
回答3:
Using simple loc and isin
cols = df.columns.tolist()
to_remove = cols[9:15] + [cols[3]]
df.loc[:, ~df.columns.isin(to_remove)]
but np.r_ is so nice I'd go with it ;)
来源：https://stackoverflow.com/questions/52245582/removing-single-and-range-of-columns-in-pandas