根据 df1 中的值在 df2 中保留一列

Question

Suppose I have two dataframes:

data = {'model':  ['B']}
df1 = pd.DataFrame (data, columns = ['model'])
df1

data = {'A':  ['0.55', '0.49', '0.47',],
        'B': ['0.48', '0.53', '0.54'],
        'C': ['0.50', '0.51', '0.45']}
df2 = pd.DataFrame (data, columns = ['A', 'B', 'C'])
df2

I would like df2 to keep only the column where the name is equal to the value in df1. So the output would be that df2 only has one column in the dataframe - 'B'

How would I do this?

Thanks

Answer 1

You can do:

import pandas as pd
data = {'model':  ['B']}
df1 = pd.DataFrame (data, columns = ['model'])
data = {'A':  ['0.55', '0.49', '0.47',],
        'B': ['0.48', '0.53', '0.54'],
        'C': ['0.50', '0.51', '0.45']}
df2 = pd.DataFrame (data, columns = ['A', 'B', 'C'])
outdf = df2[df1['model']]  # alternatively: outdf = df2[df1.model]
print(outdf)

Output:

      B
0  0.48
1  0.53
2  0.54

Explanation: we can use [] (getitem) to get certain columns from pandas.DataFrame by using iterable - in this case it is pandas.Series (column) from df1 named model .

Answer 2

Use df.isin :

In [173]: df2 = df2[df2.columns[df2.columns.isin(df1.model)]]

In [174]: df2
Out[174]: 
      B
0  0.48
1  0.53
2  0.54

OR: as per @sammywemmy's suggestion:

In [213]: df2 = df2.loc[:, df2.columns.isin(df1["model"])]

In [214]: df2
Out[214]: 
      B
0  0.48
1  0.53
2  0.54