topic Re: How do I specify values for continuous variable in SLICEBY in Graphics Programming

How do I specify values for continuous variable in SLICEBY

ljlynn97 — Mon, 20 Apr 2020 20:42:18 GMT

I'm trying to produce a slicefit type plot and want to specify values for the sliceby variable.

Outcome is a binary outcome (0/1). Variable1 and Variable2 are both continuous variables.

The code below produces exactly what I want except for one small issue: I'd like to set the values for sliceby=variable2 to be like 20, 40, 60, etc., not the 15.66, 36.36, 57.06, etc. that are being chosen automatically.

I have tried to use AT to specify values but can't get the result I want. The sliceby values are still the ones chosen automatically.

proc logistic data=my_dataset PLOTS(ONLY) = EFFECT(showobs);
	model outcome = variable1 variable2 variable1*variable2 / rsquare;
    effectplot slicefit (x=variable1 sliceby=variable2)
   	 / noobs
	;
run;

Re: How do I specify values for continuous variable in SLICEBY

Reeza — Mon, 20 Apr 2020 21:19:04 GMT

sliceby=variable2=20 40 60 80 100

Something like that should work for you.

https://blogs.sas.com/content/iml/2016/06/22/sas-effectplot-statement.html

@ljlynn97 wrote:

I'm trying to produce a slicefit type plot and want to specify values for the sliceby variable.

Outcome is a binary outcome (0/1). Variable1 and Variable2 are both continuous variables.

The code below produces exactly what I want except for one small issue: I'd like to set the values for sliceby=variable2 to be like 20, 40, 60, etc., not the 15.66, 36.36, 57.06, etc. that are being chosen automatically.

I have tried to use AT to specify values but can't get the result I want. The sliceby values are still the ones chosen automatically.
proc logistic data=my_dataset PLOTS(ONLY) = EFFECT(showobs);
	model outcome = variable1 variable2 variable1*variable2 / rsquare;
    effectplot slicefit (x=variable1 sliceby=variable2)
   	 / noobs
	;
run;

Re: How do I specify values for continuous variable in SLICEBY

ljlynn97 — Mon, 20 Apr 2020 21:26:12 GMT

Yes, thank you! I knew it had to be something simple. And thanks for the link, I'll add it to my collection.