Removing part of string from Pandas DataFrame column

Question

I have loaded a set of data to a Pandas DataFrame such as below.

test['Consultation']
Out[13]: 
0     CONSULTATION      15.00
1     CONSULTATION      10.00
2     CONSULTATION      18.00
3     CONSULTATION       0.00
4     CONSULTATION      18.00

The values are contained in the 'Consultation' column in my DataFrame .

Any idea how I can replace 'Consultation' with blank and convert the column data type to int64 or float?

My expected output is as below.

test['Consultation']
Out[13]: 
0     15.00
1     10.00
2     18.00
3      0.00
4     18.00

This is in order for myself to use DataFrame.pivot_table('Consultation',rows='Provider') to calculate the mean for my row field.

Answer 1

Why would you read the data in this way in the first place, can't you just read it into two columns? But anyway, this can be done, watch this:

In [35]:

df=pd.DataFrame({'Consultation':['CONSULTATION      15.00',
'CONSULTATION      10.00',
'CONSULTATION      18.00',
'CONSULTATION       0.00',
'CONSULTATION      18.00']})
In [36]:

import re
In [37]:

p=re.compile('[0-9.]+')
In [38]:

df['Cons']=df['Consultation'].apply(lambda x: float(p.findall(x)[0]))
In [39]:

print df
              Consultation  Cons
0  CONSULTATION      15.00    15
1  CONSULTATION      10.00    10
2  CONSULTATION      18.00    18
3  CONSULTATION       0.00     0
4  CONSULTATION      18.00    18

[5 rows x 2 columns]
In [40]:

df.dtypes
Out[40]:
Consultation     object
Cons            float64
dtype: object

In your case, you can overwrite the original df['Consultation'] by df['Consultation']=df['Consultation'].apply(lambda x: float(p.findall(x)[0]))