替换所有以数字python开头的内容

Question

I am working with ICD-9 codes for a data mining project using python and I am having trouble converting the specific codes into categories. For example, I am trying to change everything that's between 001 and 139 with 0, everything that's between 140 and 239 with 1, etc

This is what I have tried:

df = df.replace({'diag_1' : {'(1-139)' : 0, '(140-239)' : 1}})

Answer 1

You can use pd.cut to achieve this:

In [175]:
df = pd.DataFrame({'value':np.random.randint(0,20,10)})
df

Out[175]:
   value
0     12
1      2
2     10
3      5
4     19
5      2
6      8
7     14
8     12
9     16

here we set bin intervals of (0-5) (5-15), (15-20):

In [183]:    
df['new_value'] = pd.cut(df['value'], bins=[0,5,15,20], labels=[0,1,2])
df

Out[183]:
   value new_value
0     12         1
1      2         0
2     10         1
3      5         0
4     19         2
5      2         0
6      8         1
7     14         1
8     12         1
9     16         2

I think in your case the following should work:

df['diag_1']= pd.cut(df['diag_1'], [1,140,240] , labels=[1,2,3])

you can set the bins and labels dynamically using np.arange or similar

Answer 2

There is nothing wrong with an if-statement.

newvalue = 1 if oldvalues <= 139 else 2

Apply this function as a lambda expression with map .