have two df's
df1 and df2
df1 :
21 | 20 | 1 | 2 | 3 | 4 | 5 | 8 | 9 | 10
df2 :
1 | 2 | 3 | 4 | 5
abc asdf df 132 248
ban cat ball bcd aisc
how to merge two df so that i can get the desired output
output needed :
21 | 20 | 1 | 2 | 3 | 4 | 5 | 8 | 9 | 10
nan nan abc asdf df 132 248 nan nan nan
nan nan ban cat ball bcd aisc nan nan nan
You can obtain this with concat(..) [pandas-doc]:
>>> df1
Empty DataFrame
Columns: [21, 20, 1, 2, 3, 4, 5, 8, 9, 10]
Index: []
>>> df2
1 2 3 4 5
0 abc asdf df 132 248
1 ban cat ball bcd aisc
>>> pd.concat((df1, df2))
1 10 2 20 21 3 4 5 8 9
0 abc NaN asdf NaN NaN df 132 248 NaN NaN
1 ban NaN cat NaN NaN ball bcd aisc NaN NaN
This will, as the documentation says:
Concatenate pandas objects along a particular axis with optional set logic along the other axes.
Can also add a layer of hierarchical indexing on the concatenation axis, which may be useful if the labels are the same (or overlapping) on the passed axis number.
It will thus make a "union" of the column names of both dataframes, and then fill in NaNs for the columns that are missing in one of the two dataframes for the corresponding columns.
Note: The name of the columns should evidently not occur multiple times. If that happens, than it will of course error, since it is not clear how to handle such situation.
In case a column name occurs multiple times in your empty dataframe, you can resolve that with:
df1 = pd.Dataframe(columns=df1.columns.unique())
as a preprocessing step.
