Weight of Evidence and Information Value of Categorical Variable - HPBIN?

KonstantinV1 — Thu, 25 Oct 2018 08:26:10 GMT

I have a categorical variable Q1 taking the values from 1 to 4 and a binary variable Bad taking the values 1 and 0.

I want to calculate WoE and IV for the variable according to the existing categories in ascending order.

I know that HPBIN has option to calculate those statistics but it would also create other bins - i want to use the variable binned as it is. I have also tried to use the output of the Freq procedure, but it becomes inconvenient for the cases in which there is no events in some of the categories.

Re: Weight of Evidence and Information Value of Categorical Variable - HPBIN?

Ksharp — Thu, 25 Oct 2018 12:50:35 GMT

Do you want combine these into other group ? or just use its original levels ?



%let var=marital   ;


title "变量: &var";
proc sql;
create table woe_&var as
 select &var as group,
sum(good_bad='bad') as n_bad label='bad的个数',sum(good_bad='good') as n_good label='good的个数',
sum(good_bad='bad')/(select sum(good_bad='bad') from have ) as bad_dist  format=percent7.2 label='bad的占比',
sum(good_bad='good')/(select sum(good_bad='good') from have ) as good_dist  format=percent7.2 label='good的占比',
log(calculated Bad_Dist/calculated Good_Dist) as woe
from have
   group by &var
    order by woe;


select *,sum(  (Bad_Dist-Good_Dist)*woe  ) as iv
 from woe_&var ;

quit;

https://www.sas.com/content/dam/SAS/support/en/sas-global-forum-proceedings/2018/1808-2018.pdf

topic Weight of Evidence and Information Value of Categorical Variable - HPBIN? in New SAS User

Weight of Evidence and Information Value of Categorical Variable - HPBIN?

Re: Weight of Evidence and Information Value of Categorical Variable - HPBIN?