CODEWITHSUNDEEP
pythonpandasconcatenation
s_khan92
s_khan92

Reputation: 979

Merging columns in pandas?

I am merging two columns in pandas but getting a problem with "None".

my df:

id    name1    name2
1     sam      None
2     None     slic
3     brm      dres

I am using the concatenation method to combine name1 and name2

df['New']   df['name1] + df['name2']

but when i use this then all the values with None neglected... I don't know why. This is my output looks in the end

id    name1    name2     New 
3     brm      dres      brmdres

but i need this:

id    name1    name2       New
1     sam      None        sam
2     None     slic        slic
3     brm      dres        brmdres

Upvotes: 1

Views: 63

Answers (3)

Joel Leeb-du Toit
Joel Leeb-du Toit

Reputation: 171

Just a complete answer in the case where you only need to sum specific columns in the dataframe:

df = pd.DataFrame({'name1': ['sam', None, 'brm'],
                   'name2': [None, 'slic', 'dres'],
                   'name3': ['sam2', 'slic2', 'dres2']})

df['new'] = df['name1'].fillna('') + df['name2'].fillna('')

Just to explain what is happening: the fillna('') is replacing the None's with an empty string '' in a copy of the column. This means you are actually adding 'sam' to the empty string '' which gives you 'sam' again.

Upvotes: 1

Golden
Golden

Reputation: 417

A solution I like:

df['new'] = df.fillna('').sum(axis=1)

Upvotes: 0

BENY
BENY

Reputation: 323226

Try filter with replace if the None is string type, if not do fillna('')

df.filter(like='name').replace('None','').sum(1)
0        sam
1       slic
2    brmdres
dtype: object

Upvotes: 1

Related Questions