Hello
I want to calculate logistic regression model.
The dependent variable is 1/0 (Default/non-Default)
There are 2 predictors: Sex,Age (Each of them is a categorical variable).
Then I want to calculate score card for each variable and each category.
The code to calculate the scorecard is not clear for me (Someone sent it to me).
I want to ask if there is any other to calculate it?
data raw_tbl;
input sex $ age bad @@;
cards;
2 1 1 1 2 0
1 2 0 2 2 0
2 3 0 2 2 0
1 3 0 1 2 1
2 1 1 2 1 0
2 2 0 2 3 0
1 2 1 2 3 0
2 2 1 1 3 1
1 1 1 2 2 0
1 3 0 1 2 1
1 3 0 2 2 0
2 2 0 2 1 0
2 1 1 2 3 0
2 2 0 2 2 0
1 1 0 1 1 0
2 1 0 1 1 0
2 3 1 2 2 0
1 3 0 2 2 1
2 1 0 2 2 0
2 1 0 2 2 0
;
Run;
proc sql;
select sex,
count(*) as nr,
sum(bad) as nr_bads,
calculated nr_bads/calculated nr as pct_bads format=percent8.2
from raw_tbl
group by sex
;
quit;
proc sql;
select age,
count(*) as nr,
sum(bad) as nr_bads,
calculated nr_bads/calculated nr as pct_bads format=percent8.2
from raw_tbl
group by age
;
quit;
proc format;
value Dependent_Fmt
1 = 'Default'
0 = 'No_Default';
run;
proc format;
value Sex_Fmt
1 = 'Male'
2 = 'Female';
run;
proc format;
value age_Fmt
1 = 'Till_30'
2 = '30-40'
3='40+'
;
run;
ods output parameterestimates=Coef_tbl;
proc genmod data=raw_tbl namelen=60 descending ;
class sex age;
model bad=sex age/ dist=binomial link=logit type3 wald ;
store MyModel;
output out=raw_data_with_predict p=P_subscibe xbeta=logit;
/*ODS SELECT ModelANOVA;*/
run;
/**Create data set row_data_with_predict that have row data with predict colmn called :P_subscibe***/
/***Show the coefficents in a data set***/
proc plm source=MyModel;
show parameters;
run;
Data est;
set Coef_tbl;
IF Parameter in ('Intercept','Scale') then delete;
Run;
proc sort data=est(KEEP=Parameter Level1 Estimate ProbChiSq) out=est2;
by Parameter Estimate;
Run;
Data est3;
set est2 end=last;
by Parameter Estimate;
retain factor 1 max_s 0 n 0;
odds=exp(Estimate);
if first.Parameter and Estimate<1 then do;
n=n+1;
new_odds=1;
factor=1/odds;
end;
if not first.Parameter then new_odds=factor*odds;
lnodds=LOG(new_odds);
if last.Parameter then do;
factor=1;
max_s=max_s+lnodds;
end;
if last then do;
call symput('max_s',max_s);
call symput('n_var',n);
end;
drop factor;
Run;
data score_card;
set est3;
by Parameter Estimate;
_score_c=lnodds*(1000/&max_s.);
score_c=round(_score_c,1);
KEEP parameter Level1 Estimate ProbChiSqe score_c;
Run;
proc sort data=score_card;
by Parameter level1;
Run;
