Accepted Solutions
Not sure I follow all of your code. Let's just look at the most complex of those items on your screen shot.
So let's say that question on your form was named CREDIT. Then the user interface will generate these macro variables for the values shown in that picture.
%let credit0=4;
%let credit=9000=1500|2000|2300;
%let credit1=9000=1500|2000|2300;
%let credit2=11000=1800|2200|2500;
%let credit3=1500=2000|2700|3000;
%let credit4=18000=2400|3100|3500;The dataset in that last photograph does not look very useful for anything. A taller structure with fewer variables would probably work better for most purposes.
data want;
length count order tramo index months disc 8 string $100;
if not symexist('credit') then count=0;
else do;
count=1;
if not symexist('credit1') then call symputx('credit1',symget('credit'));
end;
if symexist('credit0') then count=symgetn('credit0');
do order=1 to count;
string=symget(cats('credit',order));
tramo=input(scan(string,1,'='),32.);
string=scan(string,2,'=');
months=24 ;
do index=1 to countw(string,'|');
months+12;
disc=input(scan(string,index,'|'),32.);
output;
end;
end;
drop string;
run;Results
Obs count order tramo index months disc 1 4 1 9000 1 36 1500 2 4 1 9000 2 48 2000 3 4 1 9000 3 60 2300 4 4 2 11000 1 36 1800 5 4 2 11000 2 48 2200 6 4 2 11000 3 60 2500 7 4 3 1500 1 36 2000 8 4 3 1500 2 48 2700 9 4 3 1500 3 60 3000 10 4 4 18000 1 36 2400 11 4 4 18000 2 48 3100 12 4 4 18000 3 60 3500
As I mentioned the hard-coded data _null_ works for me. I'll try your proposal and KurtBremser's as well and give you feedback.
It's a job using prompts. The context are financing campaigns for cars.
The user can choose the product and as every product has its own contract terms (36,48,60 months for "Next") and 36 to 96 by 12 for "Lineal" for which the user needs to provide discounts and credit entry levels (another user's choice, 4 or 5 credit thresholds). The more credit the final customer takes and the longer he pays off the higher discount he gets, that's the reasoning.
That makes in my case 3products X 2thresholds=6 values in the combobox.
The discount scheme for each credit threshold is controlled via a validationtext with regex.
When submitting the form I get to the job's code.
I convert these thresholds and discunts into variables because the final objective is to apply scoring code from a gradient boosting that had learnt how car model, and thresholds and discounts explain what the market values ("given that campaign 40% fall into 36 months, 38% into 48 months and 22% into 60 months X thrhesolds). At the end you see the table that has nearly all the variables the scoring code expects.
Not sure I follow all of your code. Let's just look at the most complex of those items on your screen shot.
So let's say that question on your form was named CREDIT. Then the user interface will generate these macro variables for the values shown in that picture.
%let credit0=4;
%let credit=9000=1500|2000|2300;
%let credit1=9000=1500|2000|2300;
%let credit2=11000=1800|2200|2500;
%let credit3=1500=2000|2700|3000;
%let credit4=18000=2400|3100|3500;The dataset in that last photograph does not look very useful for anything. A taller structure with fewer variables would probably work better for most purposes.
data want;
length count order tramo index months disc 8 string $100;
if not symexist('credit') then count=0;
else do;
count=1;
if not symexist('credit1') then call symputx('credit1',symget('credit'));
end;
if symexist('credit0') then count=symgetn('credit0');
do order=1 to count;
string=symget(cats('credit',order));
tramo=input(scan(string,1,'='),32.);
string=scan(string,2,'=');
months=24 ;
do index=1 to countw(string,'|');
months+12;
disc=input(scan(string,index,'|'),32.);
output;
end;
end;
drop string;
run;Results
Obs count order tramo index months disc 1 4 1 9000 1 36 1500 2 4 1 9000 2 48 2000 3 4 1 9000 3 60 2300 4 4 2 11000 1 36 1800 5 4 2 11000 2 48 2200 6 4 2 11000 3 60 2500 7 4 3 1500 1 36 2000 8 4 3 1500 2 48 2700 9 4 3 1500 3 60 3000 10 4 4 18000 1 36 2400 11 4 4 18000 2 48 3100 12 4 4 18000 3 60 3500
Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.
Click image to register for webinar
Classroom Training Available!
Select SAS Training centers are offering in-person courses. View upcoming courses for:
-
403 views
-
2 likes
-
-
