[Python] Calculate difference between rows in Pandas dataframe using conditional logic...

I am trying to use the pandas.DataFrame.diff function to calculate the difference between rows in a dataframe. The catch is I only want to calculate the difference for certain values using some simple conditional logic applied to the value in the 'data' column (i.e. if 'data' column value is 0 then set the difference to 0. Note it doesn't have to be set to 0 it could also be set as NaN).

Sample code is provided below and the table shows current results and the last column with my desired result. The 'diff_b' column almost does what I want but from outputs in the table you can see it's quite right.

import pandas as pd

# Sample DataFrame
df = pd.DataFrame({'data': [10, 20, 0, 25, 0, 0, 30, 25, 40]})

df['diff'] = df.data.diff()


Table showing results and desired result in last column

value | diff | desired_result
10 NaN NaN
20 10 10
0 -20 0
25 25 25
0 -25 0
0 0 0
30 30 30
25 -5 -5
40 15 15


I have tried getting the following code with conditional logic however this does not work

df['diff'] = df['data'].apply(lambda x: 0 if x == 0 else df.data.diff())

Continue reading...