Loop through a column of strings and NaN while avoiding NaN-CodePudding

Say I have a column called names:

names

name1
NaN
name2
NaN

Entries like name1 and name2 are of type str, and all the NaN entries are of type float, in case you needed to know.

I want to loop through this column to perform changes on the good entries while entirely leaving alone the NaN entries, as such:

for i in range(len(df["names"])):
    if df['names'][i].isnull() == False:
        # do stuff

I've tried isnull(), math.isna(), but I get errors like AttributeError: 'str' object has no attribute 'isnull' and TypeError: expected string or bytes-like object when I try to apply a lambda function instead of using a for loop.

How can I check if each row is not NaN, when all the good entries are strings?

CodePudding user response：

You can use if df['contact'][i] == df['contact'][i] to identify non-NaN values. This leverages the cool property that NaN != NaN.

Though with that said there are other probably vectorized ways to do what you are trying to do.

CodePudding user response：

If you are going to perform changes to the data of non NaN. You can consider this

df.loc[~df['name'].isna()]

This will help you to get the row of data that is not NaN.