Hello. Let's say that I have a data set with many  continuous variables (independent variables) and response variable (1/0 Default yes or no). Let's say that for specific variable X  I want to categorize this variable into categories . I have some criteria: 1-Classify to 4 or 5 categories 2-Category 1 is the worse category (The category with highest bads rate) 3-In each category be at least 50 bads (people with Y=1) 4-There must be cardinally in failure rate between groups 5-Must have at least 1% increasing in bads rate from one category to another category. 6-If there are some possible ways that meet 1-5 then choose the option with highest Information value. May anyone show a nice code that perform this? What way would you use to perform the binning based on the criteria's I mentioned?   Here is SAS code that show If the conditions I mentioned  were meet. Here can see also IV and see default rate in each category. As I said, the magic question is how to binning the categories?   PROC FORMAT ; VALUE IV_fmt low-<0.02='Very weak' 0.02 <-0.1 ='Weak' 0.1 <-0.3 ='Intermediate' 0.3 <- HIGH ='Strong' ; RUN; %macro inf_value(Raw_Data_tbl,var,Response_VAR); PROC SUMMARY DATA=&Raw_Data_tbl. (keep= &var. &Response_VAR.); VAR &Response_VAR.; OUTPUT OUT=inf0 (DROP=_TYPE_)sum=all_bad; RUN; data _null_; set inf0; good=_freq_-all_bad; call symput('all_good',good); call symput('all_bad',all_bad); run; %put &all_good; %put &all_bad; PROC SUMMARY DATA=&Raw_Data_tbl. (keep= &var. &Response_VAR.)nway; class &var.; VAR &Response_VAR.; OUTPUT OUT=inf1 (DROP=_TYPE_ rename=(_freq_=Nr_Customers))sum=Nr_bad; RUN; data inf2; set inf1; Nr_good=Nr_Customers-Nr_bad; if &all_bad. not in (0,.) then PCT_Bads=Nr_bad/&all_bad.; if &all_good. not in (0,.) then PCT_Goods=Nr_good/&all_good.; res_helkey_non=PCT_Bads/PCT_Goods; woe=log(res_helkey_non);**weight of evidance***; iv_categ=(PCT_Bads-PCT_Goods)*woe; run; PROC SUMMARY DATA= inf2; VAR iv_categ; OUTPUT OUT=inf3 (DROP=_TYPE_)sum=inf_val; RUN; data _null_; set inf3; call symputx('inf_val',inf_val); run; %put &inf_val;/***SUM IV over all categories***/ data inf4; length var $100. categ $100.; set inf2; categ=compress(&var.); drop &var.; var="&var."; P_Default=Nr_bad/Nr_Customers; label iv='information value(iv)'; run; data inf5; retain var categ Nr_Customers Nr_bad Nr_good PCT_bad PCT_Bads PCT_Goods P_Default res_helkey_non woe iv_categ ; set inf4; format PCT_Bads PCT_Goods res_helkey_non woe iv_categ P_Default percent10.3; format Nr_Customers Nr_bad Nr_good comma15. ; run; PROC SUMMARY DATA= inf5; VAR Nr_Customers Nr_bad Nr_good PCT_Bads PCT_Goods iv_categ; OUTPUT OUT=Summary_IV (DROP=_TYPE_ _freq_) sum=; RUN; Data Summary_IV_b; Retain field category ; set Summary_IV; VAR="&var."; P_Default=Nr_bad/Nr_Customers; format P_Default percent8.2; Run; Data Want; set inf5 Summary_IV_b; IF missing(categ) then IV_desc=put(iv_categ,IV_fmt.); else IV_desc=''; Run; title; proc print data=Want noobs;Run; %mend inf_value; %inf_value(Raw_Data_tbl=ttt,var=X,Response_VAR=Ind_Default)                     ... View more