Remove outliers using Pandas

remove_outliers.py

import pandas as pd
import numpy as np

def drop_outliers(df, field_name):
    distance = 1.5 * (np.nanpercentile(df[field_name], 75) - np.nanpercentile(df[field_name], 25))
    df.drop(df[df[field_name] > distance + np.nanpercentile(df[field_name], 75)].index, inplace=True)
    df.drop(df[df[field_name] < np.nanpercentile(df[field_name], 25) - distance].index, inplace=True)

if __name__ == "__main__":
  # assuming df exists and contains numeric variables

  print(df.shape)

  for column in df.select_dtypes(include=[np.number]).columns:
      drop_outliers(df, column)

  print(df.shape)