How to set K number of random column values to empty in DataFrame?-CodePudding

import pandas as pd
df = pd.DataFrame( { 'A': [1,2,3,4],
                   'B': [10,20,30,40],
                   'C': [20,40,60,80]
                  }, )



df['A'] = ''

print(df)

I want to set K number of column A's value into empty value, and these K values should be randomly selected. Those len(df)-K values of column A won't be affected. I wrote this function to generate the random row indexes, and then how to set the df's row values to empty for these indexes?

def random_rows(K=2, df):

  col_length = df.shape[1]
  row_indexes = [i for i in range(col_length)]
  if col_length<K:
     K = col_length

  selected_row_indexes = random.sample(row_indexes, K)

  return selected_row_indexes

CodePudding user response：

You can use sample to get the random rows and loc to modify them:

df.loc[df['A'].sample(n=2).index, 'A'] = '' # or whatever value you want