I use pandas to mining data. I have a DataFrame - data:
Age Sex Name
0 28 male Kirill
1 32 female Alina
2 12 female Sasha
I want to replace Sex to digit, instead male use 1, instead female - 0
I try to make in loops:
for i in range(data.Age.size()):
if data.Sex[i]=='male'
data.Sex[i]=1
else:
data.Sex[i]=0
But I get a SettingWithCopyWarning.
How I can make it right?
You can pass a dict and call map:
In [21]:
sex = {'male':1, 'female':0}
df['Sex'] = df['Sex'].map(sex)
df
Out[21]:
Age Sex Name
0 28 1 Kirill
1 32 0 Alina
2 12 0 Sasha
Or make 2 calls on the masked df:
In [25]:
df.loc[df['Sex']=='male','Sex'] = 1
df.loc[df['Sex']=='female','Sex'] = 0
df
Out[25]:
Age Sex Name
0 28 1 Kirill
1 32 0 Alina
2 12 0 Sasha
In general you should avoid looping over the df when there are vectorised solutions available, additionally it's not a good idea to mutate the container that you're iterating over as it can yield strange behaviour such as it sometimes works or it does not.
You can use the replace method, which exactly does that:
data.replace({'male': 1,
'female': 0})
Age Sex Name
0 28 1 Kirill
1 32 0 Alina
2 12 0 Sasha
or
data.replace(["male", "female"], [0, 1])
equivalent to:
data.replace(to_replace=["male", "female"], values=[0, 1])
In that case, lists must have the same length.
To add to their method, if you don't want to explicitly define the dictionary yourself you can better automate the process which comes helpful if you have multiple unique values (i.e. 5+).
import numpy as np
import pandas as pd
sex = np.sort(df['Sex'].unique()) # extract unique values and sorts them alphabetically
sex_dict = dict(enumerate(sex )) # creates a dictionary from the array above
sex_dict= dict(zip(sex_dict.values(), sex_dict.keys())) # Corrects the dictionary
df['Sex'] = df['Sex'].map(sex_dict) # maps as described in the other answers.
Again, this is more for automating the process for large unique values within the array.
Orignal DataFrame
Age Sex Name
0 28 male Kirill
1 32 female Alina
2 12 female Sasha
Final Results
Age Sex Name
0 28 1 Kirill
1 32 0 Alina
2 12 0 Sasha
