Python - 迭代列的差异

Question

Just wanted a possible solution for iterating a difference over columns. Intuition says a simple for loop should be able to solve it, however, it's been a while and I am super rusty with my syntax etc. Or a group by with a lambda function?

Problem statement.

df looks like:

Obviously working with cohort files. I want to find the difference for col(n+1) - col(n) and iterate while grouping over the cohort group.

for example for 2014-04 my result should be

Answer 1

This simple call should do the trick:

df.diff(-1, axis=1)

diff 's axis=1 will calculate a difference on the column axis. Check the doc here