CODEWITHSUNDEEP
pythonpandas
Tatik
Tatik

Reputation: 1227

How to check that pandas DataFrame field is a digit and isn't empty?

I have the following DataFrame df:

col1   col2   col3
50     dd     3
2      r      NaN
5      d      4
a      e      5

I need to calculate the mean value for selected columns cols. And then I should check if any of the values in selected rows deviate from the median value by more than 20%.

I am not sure how to tackle mixed values in a single row to make these calculations.

def test_row(x, threshold):
    if x.dtype == int or x.dtype == float:
        return x > threshold

columns = ["col1", "col3"]
for col in columns:
    threshold = df[col].median()*(20/100)
    check = df.apply(lambda x: test_row(x[col], threshold), axis=1)
    print(check.any())

However it obviously fails because if x.dtype == int or x.dtype == float does not work.

Upvotes: 0

Views: 291

Answers (1)

Felype Bastos
Felype Bastos

Reputation: 76

Your function could be:

def test_row(x, threshold):
    if isinstance(x,(int,float)) and x:
        return x > threshold

The second logic in function is just for check if x contains something, if its empty it will return False.

Upvotes: 1

Related Questions