Last answered:

07 Apr 2020

Posted on:

06 Apr 2020

0

Scorecard

I got a bunch of only nans in the lecture about creating a scorecard.            
min_score=300
max_score=850
      In [131]: min_score          
min_score=scorecard_df.groupby(['Original_names'])['Coef'].min()
min_score
    Out[131]:
Original_names
Intercept                     -1.286588
acc_now_delinq                 0.000000
addr_state                     0.000000
annual_inc                    -0.068937
dti                            0.000000
emp_length                     0.000000
grade                          0.000000
home_ownership                 0.000000
initial_list_status            0.000000
inq_last_6mths                 0.000000
int_rate                       0.000000
mths_since_earliest_cr_line    0.000000
mths_since_issue_d            -0.060627
mths_since_last_delinq         0.000000
mths_since_last_record        -0.049872
purpose                        0.000000
term                           0.000000
verification_status           -0.010562
Name: Coef, dtype: float64
In [130]: min_score_sum          
min_score_sum=scorecard_df.groupby(['Original_names'])['Coef'].min().sum()
min_score_sum
    Out[130]:
-1.4765851105053986
In [125]: max_score          
max_score=scorecard_df.groupby(['Original_names'])['Coef'].max()
max_score
    Out[125]:
Original_names
Intercept                     -1.286588
acc_now_delinq                 0.235443
addr_state                     0.524189
annual_inc                     0.577045
dti                            0.389218
emp_length                     0.126174
grade                          0.912940
home_ownership                 0.106218
initial_list_status            0.054820
inq_last_6mths                 0.694467
int_rate                       0.867411
mths_since_earliest_cr_line    0.131432
mths_since_issue_d             1.091989
mths_since_last_delinq         0.239036
mths_since_last_record         0.286678
purpose                        0.300414
term                           0.078659
verification_status            0.084432
Name: Coef, dtype: float64
In [126]: ore_sum          
max_score_sum=scorecard_df.groupby(['Original_names'])['Coef'].max().sum()
max_score_sum
    Out[126]:
5.413976291458666
In [127]: _df          
scorecard_df['Score']=scorecard_df['Coef']*(max_score-min_score)/(max_score_sum-min_score_sum)
      In [132]:        
scorecard_df.Score.unique()
    Out[132]:
array([nan])
 
1 answers ( 0 marked as helpful)
Instructor
Posted on:

07 Apr 2020

0
Hi there, Are you sure that you indexes are correct?  I would be easiest if you provide your notebook so we can examine it. Best, Iliya

Submit an answer