Pandas - Handling Nan or null entries

import pandas as pd

import numpy as np

iris = pd.read_csv("archive.ics.uci.edu/ml/machine-learning-dat..")

df = iris.copy()

df.columns = ['sl', 'sw', 'pl', 'pw', 'flower_type']

print(df)

149 rows × 5 columns

1) Method 1 - drop the entires itself

df.dropna(inplace = True)

print(df)

147 rows × 5 columns

2) Method 2 - replace the nan with a value

df.sw.fillna(df.sw.mean(), inplace = True)

df.pl.fillna(df.pl.mean(), inplace = True)

print(df)

147 rows × 5 columns

In this case, we are replacing nan values with the column's mean.

Other options are replacing with the fixed value, most recurring value, mean of the flower type and so on.

Note: Method 2 is preferred more as in most cases the no. of nan entries will be high and deleting them can lead to data loss.