Pandas: df.diff() into new column in a multi column dataframe
Question:
I have a DataFrame which is like that:
date
Col.A
Col.B
Col.C
…
2022
1
1
1
…
2023
2
2
2
…
2024
3
3
3
…
And I want to add Columns where the Diff is calculated and looks like that afterwards:
date
Col.A
Diff Col.A
Col.B
Diff Col.B
Col.C
Diff Col.C
…
2022
1
nan
1
nan
1
nan
…
2023
2
1
2
1
2
1
…
2024
3
1
3
1
3
1
…
I tried it like that, but it doesnt work
df = pd.read_excel('example.xlsx', header=0).set_index(['date])
df.diff()
How can I do that in Pandas
Answers:
df = pd.concat([df,df.diff().add_prefix('Diff ')], axis=1)
df = df.reindex(columns=sorted(df.columns, key=lambda x: (x.split(' ')[-1], x.split(' ')[0])))
df
###
date Col.A Diff Col.A Col.B Diff Col.B Col.C Diff Col.C
0 2022 1 NaN 1 NaN 1 NaN
1 2023 2 1.0 2 1.0 2 1.0
2 2024 3 1.0 3 1.0 3 1.0
I rename new columns names by DataFrame.add_suffix for correct sorting without specify all columns names:
df = pd.read_excel('example.xlsx', header=0, index_col=['date'])
df = pd.concat([df, df.diff().add_suffix(' diff')], axis=1).sort_index(axis=1)
print (df)
Col.A Col.A diff Col.B Col.B diff Col.C Col.C diff
date
2022 1 NaN 1 NaN 1 NaN
2023 2 1.0 2 1.0 2 1.0
2024 3 1.0 3 1.0 3 1.0
I have a DataFrame which is like that:
| date | Col.A | Col.B | Col.C | … |
|---|---|---|---|---|
| 2022 | 1 | 1 | 1 | … |
| 2023 | 2 | 2 | 2 | … |
| 2024 | 3 | 3 | 3 | … |
And I want to add Columns where the Diff is calculated and looks like that afterwards:
| date | Col.A | Diff Col.A | Col.B | Diff Col.B | Col.C | Diff Col.C | … |
|---|---|---|---|---|---|---|---|
| 2022 | 1 | nan | 1 | nan | 1 | nan | … |
| 2023 | 2 | 1 | 2 | 1 | 2 | 1 | … |
| 2024 | 3 | 1 | 3 | 1 | 3 | 1 | … |
I tried it like that, but it doesnt work
df = pd.read_excel('example.xlsx', header=0).set_index(['date])
df.diff()
How can I do that in Pandas
df = pd.concat([df,df.diff().add_prefix('Diff ')], axis=1)
df = df.reindex(columns=sorted(df.columns, key=lambda x: (x.split(' ')[-1], x.split(' ')[0])))
df
###
date Col.A Diff Col.A Col.B Diff Col.B Col.C Diff Col.C
0 2022 1 NaN 1 NaN 1 NaN
1 2023 2 1.0 2 1.0 2 1.0
2 2024 3 1.0 3 1.0 3 1.0
I rename new columns names by DataFrame.add_suffix for correct sorting without specify all columns names:
df = pd.read_excel('example.xlsx', header=0, index_col=['date'])
df = pd.concat([df, df.diff().add_suffix(' diff')], axis=1).sort_index(axis=1)
print (df)
Col.A Col.A diff Col.B Col.B diff Col.C Col.C diff
date
2022 1 NaN 1 NaN 1 NaN
2023 2 1.0 2 1.0 2 1.0
2024 3 1.0 3 1.0 3 1.0