CODEWITHSUNDEEP
pythonpandasdataframe
Kristoffer Tølbøll
Kristoffer Tølbøll

Reputation: 3357

Dataframe can't fill NaN values in pandas

I have a DataFrame with some NaN values, that I want to fit with the mean of my DataFrame

data.fillna(data.mean())

afterwards, I check to see if there are any NaN or null values left

 print('is nan', data.isna().values.any()) #prints True
 print('is null', data.isnull().values.any()) #prints True

before prints to be true, and I don't understand since the NaN values should be filled.

Am I missing something?

Upvotes: 1

Views: 1034

Answers (2)

ewalel
ewalel

Reputation: 2096

This is a sample reproducible example.

import pandas as pd
# generate sample dataframe some filled with NaN
data = pd.DataFrame([[1, 2], [np.nan, 4], [5, 6], [7, np.nan]], columns=["A", "B"])
# apply per-column and substitue the mean of that columns 
data.apply(lambda x: x.fillna(x.mean()),axis=1)

Upvotes: 1

petezurich
petezurich

Reputation: 10224

Try this to change your dataframe in place:

data.fillna(data.mean(), inplace=True)

Or do this to create a new dataframe with NaNs replaced by data.mean():

data_new = data.fillna(data.mean())

Upvotes: 2

Related Questions