Myujth

Suppose I have a dataframe with columns as:
JobTitle, Age, Stats and Date. Goal is to group it by JobTitle and Age and apply the Aggregation functions to Stats and Date. Stats column will be read from a config file. If it is empty, the default will be 'mean', else it will take the user defined number for percentile.

This is what I have done:

import pandas as pd
import numpy as np
aggregate_dict='Stats':'Mean', 'Date':'min'

for i in range(0,df.shape[0]):
 if df.Stats[i]:
 temp_StatsName = df.Stats[i]
 aggregate_dict='Stats':'percentile('+temp_StatsName+')', 'Date':'min'


df_final=df.groupby(['JobTitle','Age']).agg(aggregate_dict).reset_index()

Besides this, I have also tried creating my own percentile function as so and used it in the aggregate_dict definition but was not successful:

def percentile(n):
 def percentile_(x):
 return np.percentile(x, n)
 percentile_.__name__ = 'percentile_%s' % n
 return percentile_

If you may have a suggestion how to implement user defined percentiles when the Stats column is not empty, that would be very helpful. My code may not be ideal, I am relatively new in Python. Thanks!