Setting constraints for multiple linear regression coefficients

[Mbs270]

I plan on using proc reg to do multiple linear regression. I need to set a constraint on the coefficients SAS arrives at.

If Y = C1X1 + C2X2 + C3X3 + ..., I need the constraint that C1 > C2 > C3 ....

I've read about the restrict statement and proc calis but am not sure which to use. Thanks in advance.

[PaigeMiller]

There is no such constraint C1 > C2 > C3 ... possible in PROC REG. The only constraints possible in PROC REG are to see if a linear combination of regression coefficients is equal to a specific value (not greater than a specific value).

I'm not very familiar with PROC CALIS, but it appears that such a constraint is possible there.

[Mbs270]

Can anyone offer insights on using proc calis to do what I want? Cannot find syntax examples for what I want to do on the internet.

[PaigeMiller]

Plenty of examples in the PROC CALIS documentation

https://documentation.sas.com/?cdcId=pgmmvacdc&cdcVersion=9.4&docsetId=statug&docsetTarget=statug_ca...

[Ksharp]

I remembered @Rick_SAS has written a blog about it. Check PROC HPGLM, GENMOD , HPREG , HPGLMSELECT

[Rick_SAS]

I believe you can use the RESTRICT statement in PROC HPGENSELECT. See the SAS Note for this situation.