nicopalacios Tracker

Re: Error in SAS Miner, node Score

nicopalacios — Mon, 24 Apr 2017 15:48:42 GMT

Número de variables con atributos diferentes: 2

Listado de variables comunes con atributos diferentes

Variable Conjunto de datos / Tipo / Longitud /= Formato / F.lect. Etiqueta

Pagos_t

ANDA.TABLAFINALCOMPLETA / Num / 8 / BEST8. / BEST8

OCT2016.TABLAOCTUBRE2016 / Num / 8 / BEST9. / BEST9.

Pagos_t1

ANDA.TABLAFINALCOMPLETA / Num / 8 / BEST8. / BEST8.
OCT2016.TABLAOCTUBRE2016 / Num / 8 / BEST9./ BEST9.

What is Different attributes?? The type, lengh, and format is equal!

Re: Error in SAS Miner, node Score

nicopalacios — Mon, 24 Apr 2017 15:13:44 GMT

Yes, i have EM. How i run this? I run a SAS program? I put only PROC COMPARE?

Re: Error in SAS Miner, node Score

nicopalacios — Fri, 21 Apr 2017 20:27:49 GMT

Thanks Reeza.

Where i run PROC COMPARE? In a SAS Code? In Score node?

Error in SAS Miner, node Score

nicopalacios — Fri, 21 Apr 2017 19:01:29 GMT

Hi everyone! I have a problem..

I built a Credit Scoring model, all is right. But when i use a dataset for score the model, do not run.

I attach the flow:

And this is the log error:

10870 proc freq data=EMWS1.Score_VariableSet noprint;
10871 table ROLE*LEVEL/out=WORK.ScoreMETA;
10872 run;

NOTE: There were 2 observations read from the data set EMWS1.SCORE_VARIABLESET.
NOTE: The data set WORK.SCOREMETA has 2 observations and 4 variables.
NOTE: PROCEDIMIENTO FREQ used (Total process time):
real time 0.01 seconds
cpu time 0.01 seconds

10873 proc print data=WORK.ScoreMETA label noobs;
10874 var ROLE LEVEL COUNT;
10875 label ROLE = "%sysfunc(sasmsg(sashelp.dmine, meta_role_vlabel, NOQUOTE))" LEVEL = "%sysfunc(sasmsg(sashelp.dmine, meta_level_vlabel, NOQUOTE))" COUNT = "%sysfunc(sasmsg(sashelp.dmine, rpt_count_vlabel, NOQUOTE))";
10876 title9 ' ';
10877 title10 "%sysfunc(sasmsg(sashelp.dmine, rpt_varSummary_title , NOQUOTE))";
10878 run;

NOTE: There were 2 observations read from the data set WORK.SCOREMETA.
NOTE: The PROCEDURE PRINT printed page 1.
NOTE: PROCEDIMIENTO PRINT used (Total process time):
real time 0.01 seconds
cpu time 0.01 seconds

10879 title10;

10880 %let groupid =;
10881 %let endGroupid =;
10882 %let prescoreid =;
10883 %let idsTable =;
10884 %let hpdmTable =;
10885 %let lasthptm =;
10886 data _null_;
10887 set EMWS1.Scorecard_EMINFO;
10888 where upcase(key) in('LASTHPTM', 'IDSTABLE', 'HPDMSAMPLE', 'PRESCORECODE','ENDGROUP', 'GROUPINFO', 'BOOSTINFO', 'BAGINFO', 'TARGETINFO', 'INDEXINFO');
10889 select(upcase(key));
10890 when('LASTHPTM') call symput('lasthptm', DATA);
10891 when('IDSTABLE') call symput('idsTable', DATA);
10892 when('HPDMSAMPLE') call symput('hpdmSample', DATA);
10893 when('PRESCORECODE') call symput('prescoreId', DATA);
10894 when('ENDGROUP') call symput('endGroupId', DATA);
10895 otherwise call symput('groupId', DATA);
10896 end;
10897 run;

NOTE: There were 1 observations read from the data set EMWS1.SCORECARD_EMINFO.
WHERE UPCASE(key) in ('BAGINFO', 'BOOSTINFO', 'ENDGROUP', 'GROUPINFO', 'HPDMSAMPLE', 'IDSTABLE', 'INDEXINFO', 'LASTHPTM', 'PRESCORECODE', 'TARGETINFO');
NOTE: DATA statement used (Total process time):
real time 0.01 seconds
cpu time 0.01 seconds

10898 %let lastModelid =;
10899 data _null_;
10900 set EMWS1.Scorecard_EMINFO;
10901 where upcase(key) = 'MODEL' and TARGET = "Objetivo";
10902 call symput('lastModelID', DATA);
10903 run;

NOTE: There were 1 observations read from the data set EMWS1.SCORECARD_EMINFO.
WHERE (UPCASE(key)='MODEL') and (TARGET='Objetivo');
NOTE: DATA statement used (Total process time):
real time 0.00 seconds
cpu time 0.00 seconds

Executing SASHELP.EMCORE.EMINFOITERATOR.SCL _INIT >>
Executing SASHELP.EMCORE.EMINFOITERATOR.SCL setMetaData >>
NOTE: There were 1 observations read from the data set EMWS1.SCORECARD_EMINFO.
WHERE (TARGET='Objetivo') and (KEY='MODEL');
NOTE: The data set WORK.EMINFO42DYEDYL has 1 observations and 3 variables.
Executing SASHELP.EMCORE.EMINFOITERATOR.SCL NUMELEMENTS >>
Executing SASHELP.EMCORE.EMINFOITERATOR.SCL _term >>
Executing SASHELP.EMCORE.EMINFOITERATOR.SCL _INIT >>
Executing SASHELP.EMCORE.EMINFOITERATOR.SCL setMetaData >>
NOTE: There were 1 observations read from the data set EMWS1.SCORECARD_EMINFO.
WHERE (TARGET='Objetivo') and (KEY='DECMETA');
NOTE: The data set WORK.EMINFO42DWO6FN has 1 observations and 3 variables.
Executing SASHELP.EMCORE.EMINFOITERATOR.SCL next >>
Executing SASHELP.EMCORE.EMINFOITERATOR.SCL next >>
Executing SASHELP.EMCORE.EMINFOITERATOR.SCL _term >>
NOTE: This SAS session is using a registry in WORK. All changes will be lost at the end of this session.
WARNING: Variable CatBCUt1 has already been defined as numeric.
12029 _UFormat = put(CatBCUt1,$CHAR7.);
-------
48
WARNING: Variable Ramo has already been defined as numeric.
12227 _UFormat = put(Ramo,$CHAR50.);
--------
48
WARNING: Variable Sector has already been defined as numeric.
12329 _UFormat = put(Sector,$CHAR24.);
--------
48
ERROR 48-59: formato CHAR no se he encontrado o no se puede cargar.

*------------------------------------------------------------*
*
* ERROR: Se ha encontrado un error de tiempo de ejecución. El error del sistema es -S-.
* Reporte los problemas sin resolver al departamento de Soporte técnico.
*
*------------------------------------------------------------*

Can anybody help me?

how to balance models with SAS Miner?

nicopalacios — Tue, 24 Nov 2015 16:30:49 GMT

Hello everyone!
I'm doing a model socring credit but this is not well ordered. The Hosmer-Lemeshow did not return good results. There are any way that the model better in SAS Miner?

Thanks!

cutoff modifying and affecting statistical?

nicopalacios — Mon, 17 Aug 2015 15:41:14 GMT

Hello everyone, I have a question..

I'm building a credit scoring model in SAS Miner. My problem is that the statistics that throws the "Scorecard" node are based on a cutoff of 0.50. I need to use another cutoff, but the statistics are based on this new cutoff.

For example, the percentage of correctly predicted predictions that looks at the results of the "Scorecard", are based on the new cutoff.

Is this possible?

Regards!

Re: How I can change the cutoff?

nicopalacios — Tue, 07 Jul 2015 13:34:30 GMT

excellent!

thank you

Re: How I can change the cutoff?

nicopalacios — Tue, 07 Jul 2015 12:36:34 GMT

Hi Miguel, thanks for your answer.

I can not find where to change the cutoff. I tried in modificarle the "User imput," but it turns out to see the "data exported" continues using 0.50 as the cutoff. And in the Score node can not find an option to change the cutoff. What would be the correct procedure? Scorecard > Cutoff > Score?

How I can change the cutoff?

nicopalacios — Mon, 06 Jul 2015 14:35:14 GMT

Hello everyone!

I have a problem. I'm doing a credit scoring model, and the cutoff of probability is 0.50. How should I do to change the cutoff?

How does the Hosmer-Lemeshow test??

nicopalacios — Fri, 03 Jul 2015 19:26:34 GMT

Hello everyone!

I need to know how to see the Hosmer-Lemeshow test in SAS Miner. I'm building a model of credit scoring, and I have Scorecard output. Where I can see such a test?

thanks!

Re: how to make a correlation analysis between the variables?

nicopalacios — Tue, 30 Jun 2015 04:04:05 GMT

Miguel, tus respuestas siempre son excelentes! Te agradezco

p,d: tu español es muy bueno, no te preocupes.

Saludos y gracias

Re: how to make a correlation analysis between the variables?

nicopalacios — Mon, 29 Jun 2015 13:51:20 GMT

Hello Miguel! thanks for your answer..

I have not been able to resolve. There is a node or output table that shows me directly the correlation between the variables?

Te escribo esto en español ya que veo que sos latino igual que yo. El problema que estoy teniendo es que estoy construyendo un modelo de credit scoring, pero SAS no me toma como variables significativas unas X que para nosotros son FUNDAMENTALES, entonces lo que queremos es ver la correlación entre las variables, para así poder excluir las variables que tengan una correlación muy alta con las X que consideramos fundamentales, por mas que la significación global disminuya un poco.

how to make a correlation analysis between the variables?

nicopalacios — Fri, 26 Jun 2015 20:25:25 GMT

Hello everyone!

I need to do in the Miner a correlation analysis between variables. how I can do this?

Regards

how do to change the language SAS Miner and SAS Guide?

nicopalacios — Fri, 12 Jun 2015 19:37:43 GMT

Hello everyone!

My question is simply, how do to change the language of SAS english to spanish?

Thanks!

Re: SAS® Enterprise Miner™ can build credit scorecards

nicopalacios — Fri, 12 Jun 2015 19:34:11 GMT

thanks Anna!

Re: Tip: How to build a scorecard using Credit Scoring for SAS® Enterprise Miner™

nicopalacios — Fri, 12 Jun 2015 19:10:10 GMT

Miguel, the model was great! We could do what you told me. We would like to make one last question and thus not bother you anymore. At the time of selecting the observations to 1 or 0 by prediction probability, SAS uses as a cutoff 0.50. How I can change this value cutoff?

Re: Tip: How to build a scorecard using Credit Scoring for SAS® Enterprise Miner™

nicopalacios — Thu, 11 Jun 2015 18:53:45 GMT

Thank you very much to both of you, it was very helpful to us. By connecting the new data source to the interactive grouping and running it, I an error shows up (in the photo), perhaps it is because some name differences between the two databases.

Now I'll try what you recommend maybe I will write again.

I thank you again

Re: Tip: How to build a scorecard using Credit Scoring for SAS® Enterprise Miner™

nicopalacios — Thu, 11 Jun 2015 17:53:16 GMT

Miguel thanks for your help! Then I have to connect my new set of data to the same interactive grouping model? Do I have to add any other node?

Re: Tip: How to build a scorecard using Credit Scoring for SAS® Enterprise Miner™

nicopalacios — Thu, 11 Jun 2015 15:48:27 GMT

Mendy thank you very much! You're being a great help for us

The new data source I put "Role: test" and connect to the scorecard, but I get an error, as shown in the photo.

Previously had connected the new data source to "interactive grouping" and also appeared the same sign error. This may be because the columns of data sources, although representing the same variables do not have exactly the same name?

And one last question to not disturb more, the new data source would have to connect to the same "interactive grouping" model, or a new "interactive grouping" and that this is connected to the scorecard?

greatly appreciate your help

Re: Tip: How to build a scorecard using Credit Scoring for SAS® Enterprise Miner™

nicopalacios — Tue, 09 Jun 2015 15:25:45 GMT

I do not understand the process well. I have the scorecard node, but I don't know how to connect the new data source (that new data source in which I want to validate the prediction %).Is this as in the attached image or am I wrong?