根据存储在不同数据框中的行和列标签从熊猫数据框中选择值

Question

I have two dataframes: the first contains rating and tenor per ID

df1 = 
ID     RATING     TENOR
1234   BBB        2.0
2345   BB+        1.5
3456   BBB-       1.0

The ID is the index column.

In the second dataframe df2 I have a matrix of rating as index and tenor as column names.

         0.5       1.0       1.5       2.0  
BBB+  0.000750  0.001500  0.002651  0.003800  
BBB   0.001201  0.002400  0.003751  0.005100  
BBB-  0.001401  0.002800  0.006256  0.009700  
BB+   0.002854  0.005700  0.009106  0.012500  
BB    0.004058  0.008100  0.014470  0.020800  
BB-   0.006723  0.013400  0.024108  0.034700  
B+    0.011567  0.023000  0.043005  0.062600  
B     0.033874  0.066600  0.086257  0.105500  
B-    0.048475  0.094600  0.124649  0.153700  

now I would like to select the value that corresponds to the rating and the tenor of each id in df1 and add it to a new column in df1. So the result should look like:

ID     RATING     TENOR   PD  
1234   BBB        2.0     0.005100      
2345   BB+        1.5     0.009106  
3456   BBB-       1.0     0.002800

I tried
df1['PD'] = df2.loc[df1.RATING, df1.TENOR].values

but this doesn't work. Thanks for your help.

Answer 1

Use lookup :

df1['PD'] = df2.lookup(df1.RATING,df1.TENOR.astype(str))

Output:

     RATING  TENOR        PD
ID                          
1234    BBB    2.0  0.005100
2345    BB+    1.5  0.009106
3456   BBB-    1.0  0.002800

Note: I had to use astype because my dtypes didn't match in my df2.columns and df1.TENOR. You'll have to check this you might not need to astype depending on your datatypes.