I have dataframe with 3 columns (id, parentId, version) and I need to add new column parentVersion (version from parentid) :
| id | parentid | version | parentVersion |
|---|---|---|---|
| 1 | 2 | 1.1 | 2.1 |
| 2 | 2 | 2.1 | 2.1 |
Add a comment |
2 Answers
Let's try creating a mapper out of the id and version columns with set_index then calling Series.map on the parentid column:
df['parentVersion'] = df['parentid'].map(df.set_index('id')['version']) df:
id parentid version parentVersion 0 1 2 1.1 2.1 1 2 2 2.1 2.1 The mapper:
df.set_index('id')['version'] Is a Series of values which associates ids with version:
id 1 1.1 2 2.1 Name: version, dtype: float64 Then map will match the value in the index of this series (from the parentid column) and replace it will the corresponding value.
Alternatively a self merge on id and parentid then rename and filter out the unneeded columns:
df = df.merge( df, how='left', left_on='parentid', right_on='id', suffixes=('', '_y') ).rename(columns={'version_y': 'parentVersion'}).filter(regex='.*(?<!_y)$') df:
id parentid version parentVersion 0 1 2 1.1 2.1 1 2 2 2.1 2.1 
Comments
Define a dictionary containing the data : data = {"id":[1,2], "parentid":[2,2], "version":[1.1,2.1], }
Convert the dictionary into DataFrame:df = pd.DataFrame(data)
Declare a list that is to be converted into a column: parentVersion= [2.1, 2.1]
Using 'parentVersion' as the column name: df['parentVersion'] = parentVersion
Observe the result: print(df)
