CODEWITHSUNDEEP
python-3.xpandasdataframe
Codegator
Codegator

Reputation: 637

Create new column in python 3 (pandas) dataframe based on value in other column

I have a pandas dataframe where I need to create new columns based on values from other columns in dataframe. Here is the dataframe

person city state country
A Chicago Illinois USA
B Phoenix Arizona USA
C San Diego California USA

I want to create two new columns based on the values in state

  1. Create new column df["city-north"] = df['city'] where state = "Illinois"
  2. Create new column df["city-south"] = df['city'] where state is not equal to "Illinois"

I tried

df.loc[((df['state'] == 'Illinois')), 'city-north'] = df['city']
df.loc[((df['state'] != 'Illinois')), 'city-south'] = df['city']

But second line of code where not equal to condition is there does not create 'city-south' column. Please help

Upvotes: 1

Views: 1095

Answers (2)

Julius Sechang Mboli
Julius Sechang Mboli

Reputation: 60

I think your code should work fine.

df.loc[df['state'] == 'Illinois', 'city-north'] = df['city']
df.loc[df['state'] != 'Illinois', 'city-south'] = df['city']

I even tried it with multiple conditions and new inputs that were not in the original dataframe and still worked as follows:

df.loc[(df['state'] == 'Illinois') & (df['city'] =='Chicago'), 'city-north'] = 'New Entry']
df.loc[(df['state'] != 'Illinois') & (df['city'] =='Phoenix '), 'city-south'] = 'Another new entry'

Kindly note that the second part of my answer is to show more clarity with multiple conditions and it still worked. It can be used to add new columns with new entries.

Upvotes: 0

jezrael
jezrael

Reputation: 862441

For me working well, if no match conditions are created missing values:

df.loc[df['state'] == 'Illinois', 'city-north'] = df['city']
df.loc[df['state'] != 'Illinois', 'city-south'] = df['city']

print (df)
  person       city       state country city-north city-south
0      A    Chicago    Illinois     USA    Chicago        NaN
1      B    Phoenix     Arizona     USA        NaN    Phoenix
2      C  San Diego  California     USA        NaN  San Diego

If need empty values strings for not matched rows:

df['city-north'] = np.where(df['state'] == 'Illinois', df['city'], '')
df['city-south'] = np.where(df['state'] != 'Illinois', df['city'], '')

print (df)
  person       city       state country city-north city-south
0      A    Chicago    Illinois     USA    Chicago           
1      B    Phoenix     Arizona     USA               Phoenix
2      C  San Diego  California     USA             San Diego

Upvotes: 1

Related Questions