Conditional Logit in PROC MODEL to enable use of PDL macro

[PamG]

I need to use distributed lag model (DLM) on a binary outcome.  There is PDLREG but it can be applied to continuous outcomes. So it seems like the only way I can do this is using PROC MODEL and use %pdl macro.  However I am using a 1:5 matched data and need to use conditional logit.  How does one implement a conditional logit model in PROC MODEL?  I am aware that conditional logit model can be implemented in PROC PHREG or PROC LOGISTIC.  My need to use PROC MODEL is that I can use the pdl macro.

A sample code is given below for a continuous dependent variable.  I would like a similar code for conditional logit that would take into account the matched sets.  

proc model data=pdl;
   parms int;                  /* declare the intercept parameter */
   %pdl( xpdl, 6, 4 )          /* declare the lag distribution */
   y = int + %pdl( xpdl, x );  /* define the model equation */
   fit y / list;               /* estimate the parameters */
run;

Ideally I need to use DLNM model but unfortuantely SAS does not have a procedure in place for this .  Now I need to use DLM but seems like I have hit a roadblock with that too.  

[sbxkoenk]

Hello,

Cannot answer your question, but moving it to more appropriate board ("Forecasting and Econometrics" - board).

SAS/ETS User's Guide
The MODEL Procedure
Distributed Lag Models and the %PDL Macro
https://go.documentation.sas.com/doc/en/pgmsascdc/9.4_3.5/etsug/etsug_model_sect170.htm

Koen

[PamG]

Thanks.  But it seems to me that this is also a SAS programming question of how to use PROC MODEL to implement a conditional logit model.

[sbxkoenk]

While most (the big majority) of the procedures are "pre-written" programs that are built in, some procedures do indeed allow for additional programming within the boundaries of the initial "PROC" and the ending "run;" or "run; quit;" keywords.

Think about PROC NLMIXED, PROC MCMC, PROC MODEL, ...

This is indeed programming, but it still requires a lot of statistical / econometrical knowledge.

The programming board is about data management, data step programming, SQL, macro language, global statements and so on ...

Br, Koen

[PamG · Accepted solution]

An update, Almon lags can easily be created in data step. I will be using these derived variables in conditional logit using PROC LOGISTISTIC and then back-tranform the effects to the original lagged variables.   I wish there was an easier way to do it.