重命名数据框中的列名，该列名具有与旧名称相似的字符串

Question

I want to rename a column name in dataframe. Data is coming on fly and column name may contain some non english characters as well. So I am looking for something like:

df = df.rename(columns={'string include (oldName1)': 'newName1', 'string include (oldName2)': 'newName2'}

Present rename function in Pandas is:

df = df.rename(columns={'oldName1': 'newName1', 'oldName2': 'newName2'})

This function require exact column name 'oldName1'. I want to rename column even if old name includes some string like/in 'oldName1'.

Answer 1

You can use df.columns.str.contains :

df = pd.DataFrame({'oldName1': [1,1], 'zzzoldName2': [2,2], 'someoldName3': [3,3]})

converters = {'oldName1': 'newName1','oldName2': 'newName2','oldName3': 'newName3'}
conversion_map = {df.columns[df.columns.str.contains(old)][0]: new for old, new in converters.items()}

df.rename(columns=conversion_map)

>>    newName1  newName3  newName2
0         1         3         2
1         1         3         2

Note that the [0] will return the first match, so you may want to check conversion_map to see how the name is changed. If a name cannot be found this way it will raise an IndexError