Why does PROC ADAPTIVEREG not output certain options such as COOKD?

[SASSYstats]

Hello everyone, this is my first question to the forum.

I have a data set of about 45,000 observations. The following variables are character: Room_Type, Cluster_value, min_nights. All other variables are numeric. The question has to do with the OUTPUT statement. I am able to output the results into the capstn library just fine except for when I include the option COOKD=CD (as shown below). Everything works except for that. COOKD lights up blue and from what I can see in the documentation there is no reason it shouldn't be available. I get error number 73 EXPECTING AN =. I have found the same issue to be true when I try outputting studentized residuals. 

proc adaptivereg data=Training;
	class Room_Type Cluster_value min_nights;
	model price_rank=min_nights Room_Type Cluster_value avail_rank reviews_rank / additive;
	output predicted=Pred residual=R cookd=CD out=capstn.addaptive_mod;
	title "Regression Analysis for Price of Listing by Rank";
run;

title;

 If anyone has any idea this has been stumping me for some time now. I am using the latest version of SAS University Edition on Windows 10.

Thanks Everyone!

[Reeza · Accepted solution]

I don't see COOKD listed as an option in adaptivereg, where are you seeing it as a valid option?

https://documentation.sas.com/?docsetId=statug&docsetVersion=15.1&docsetTarget=statug_adaptivereg_sy...

---

@SASSYstats wrote:

Hello everyone, this is my first question to the forum.

I have a data set of about 45,000 observations. The following variables are character: Room_Type, Cluster_value, min_nights. All other variables are numeric. The question has to do with the OUTPUT statement. I am able to output the results into the capstn library just fine except for when I include the option COOKD=CD (as shown below). Everything works except for that. COOKD lights up blue and from what I can see in the documentation there is no reason it shouldn't be available. I get error number 73 EXPECTING AN =. I have found the same issue to be true when I try outputting studentized residuals. 

proc adaptivereg data=Training;
	class Room_Type Cluster_value min_nights;
	model price_rank=min_nights Room_Type Cluster_value avail_rank reviews_rank / additive;
	output predicted=Pred residual=R cookd=CD out=capstn.addaptive_mod;
	title "Regression Analysis for Price of Listing by Rank";
run;

title;

 If anyone has any idea this has been stumping me for some time now. I am using the latest version of SAS University Edition on Windows 10.

Thanks Everyone!

---

[SASSYstats]

Well this may be a newbie question but are you saying that just because an option is listed after a statement on Univ. Ed., and you can retrieve info on it, and it lights up blue, it isn't necessarily supported?

[Reeza]

---

@SASSYstats wrote:
Well this may be a newbie question but are you saying that just because an option is listed after a statement on Univ. Ed., and you can retrieve info on it, and it lights up blue, it isn't necessarily supported?

---

I can't speak to that portion of SAS unfortunately. I don't use the coding popups as I find them obtrusive. It's possible it's just pulling from a generic OUTPUT statement or it's possible it's a bug. Perhaps worth a quick conversation with SAS tech support. I do know there are definitely issues with words being in blue/red and that's not a 100% an indicator of correct or incorrect code. Sorry, not super helpful but hope it solves your initial question. 

[Ksharp]

PROC ADAPTIVEREG is non-parameter method .
It should not output some stat like COOKD .
@Rick_SAS may explain it further .

[Rick_SAS]

I can, but Reeza already gave the correct answer: The documentation for PROC ADAPTIVEREG shows that the COOKSD= option is not supported. 

As KSharp says, it is not supported because that statistic applies only to linear regression models. To learn more about Cook's D statistic, see "Influential observations in a linear regression model: The DFFITS and Cook's D statistics."

[SASSYstats]

Thanks everyone. In hindsight several beginner mistakes on my part. 

Thanks for the article @Rick_SAS  very informative.