CODEWITHSUNDEEP
pythonpandasmissing-datanonetype
singularity2047
singularity2047

Reputation: 1071

Change None value in a column to string 'None' in Python

Some columns in my data set have missing values that are represented as None (Nonetype, not a string). Some other missing values are represented as 'N/A' or 'No'. I want to be able to handle these missing values in below method.

df.loc[df.col1.isin('None', 'Yes', 'No'), col1] = 'N/A'

Now my problem is, None is a value not a string and so I can not use none as 'None'. I have read somewhere we can convert that none value to a string 'None'.

Can anyone kindly give me any clue how to go about it ?

Note 1:

Just for clarity of explanation if I run below code:

df.col1.unique()

I get this output:

array([None, 'No', 'Yes'], dtype=object)

Note 2:

I know I can handle missing or None value with isnull() but in this case I need to use .isin() method

Sample dataframe:

f = {'name': ['john', 'tom', None, 'rock', 'dick'], 'DoB': [None, '01/02/2012', '11/22/2014', '11/22/2014', '09/25/2016'], 'Address': ['NY', 'NJ', 'PA', 'NY', None]}
df1 = pd.DataFrame(data = f)

When you run below code you will see None as a value.

df1.Address.unique()
output: array(['NY', 'NJ', 'PA', None], dtype=object)

I want the None to be displayed as 'None'

Upvotes: 0

Views: 3476

Answers (2)

Bryce Ramgovind
Bryce Ramgovind

Reputation: 3257

There is a different between a null/None and 'None'. So you can change your original statement to 

df.loc[df.col1.isin([None, 'Yes', 'No']), col1] = 'N/A'

That is, take out the apostrophes for None

Or you can first find all the indices where a null's or none's exist and then select all those rows based on the index. And then you can use your original statement. 

df["col1"].loc[df["col1"].isnull()] = 'None'

Upvotes: 1

PdevG
PdevG

Reputation: 3677

Create an example df:

df = pd.DataFrame({"A": [None, 'Yes', 'No', 1, 3, 5]})

which looks like:

     A
0  None
1   Yes
2    No
3     1
4     3
5     5

Replace your 'None' by None and make the to be replaced arguments a list (that's how isin works):

df.loc[df.A.isin([None, 'Yes', 'No']), 'A'] = 'N/A'

which returns:

     A
0  N/A
1  N/A
2  N/A
3    1
4    3
5    5

Upvotes: 0

Related Questions