How to analyze data from one df to every n data?-CodePudding

I'm currently working with a pandas df where I need to perform an analysis for each 96 datas so far I am doing the process manually usind .loc function :

df_partial1 = df.loc[0:95]
df_partial2  = df.loc[96:191]

And so on . For each df_partial I'm applying a function called bill. I'd like to automate this process in order to make it more efficient. Something like this:

a = [] # empty list 
for df_partial (each 96 points) in the df:
        a = bill(df_partial)
        a = a.append(a) # A list with each result of the function for each df_partial

Can anyone help me with this?

CodePudding user response：

You can try to have a look at window functions with pandas. I believe the first example would help you: https://pandas.pydata.org/pandas-docs/stable/user_guide/window.html

It also have an apply method which allows you to directly apply the function, instead of iterating through the windows and call your function.

CodePudding user response：

Use groupby and a custom group:

import numpy as np

group = np.arange(len(df))//96
# group 0 is 0:95, group 1 is 96:191, etc.

a = [bill(d) for _,d in df.groupby(group)]