Counting Nan and null values in a pandas dataframe

Question

Imagine we have a CSV file called.csv data:

col1    col2    col3   col4
1        2        3        4
5       6         7        8
9      10        11       12
13     14        15    
33    44

import numpy as np
import pandas as pd

po = pd.read_csv('/dados.csv')

My goal is to better understand how to identify Nan/null data in a dataset.

Questions:

1.How to count how many 'Nan' data are in the above dataset?

2.How to count how many null data there are in the above dataset?

3.How to count how many data NOT 'Nan' exist in the above dataset?

4.How to count how many non-null data there are in the above dataset?

And the same questions above but per column?

I tried, for example,:

po[po['col4'].isna()].count()

thinking of accounts how many 'Nan' exist in column 4, but the answer was:

col1    2
col2    2
col3    1
col4    0
dtype: int64

What is wrong? How to answer the above questions?

Answer 1

What is wrong?

The function count() does non-zero data counting (for each column or row), the correct use of it is:

• Non-zero data count for all columns

  print(po.count())
  

  the exit will be:

  col1    5
  col2    5
  col3    4
  col4    3
  dtype: int64

• Non-zero data count for a specific column

  print(po.col4.count())
  

  the exit will be:

  3

See working on repl.it

To count missing data, you can use the function isna() or the function isnull()

• Missing data count from all columns

  # isna
  print(po.isna().sum())
  
  # isnull
  print(po.isnull().sum())
  

  the output of both will be:

  col1    0
  col2    0
  col3    1
  col4    2
  dtype: int64

• Count of missing data from a specific column

  # isna
  print(po.col4.isna().sum())
  
  # isnull
  print(po.col4.isnull().sum())
  

  the output of both will be:

  2

See working on repl.it

References:

• pandas.DataFrame.Count
• pandas.DataFrame.Isna
• pandas.DataFrame.isnull

Answer 2

1 and 2: how much Nan data is in the above dataset -> po.Isna(). sum(). sum()