n by n matrix of type 3 tests of fixed efetcs

[Nicolai70]

Hi.

I have used this code to generate a n by n run and obtain 19 times 6 results:

Is it possible to create a matrix output of the type 3 tests only in the form of "interval" by "ud_size" and highlight the significant (on a 0.05 level) effects? I am ony interested in the P values.

Or, should i use another method?

Nicoai

proc sort data=ARTICLE2.TEMP_5_60_SPAT_5_95 out=Work.TempDataSorted;
	by interval ud_size ;
run;

proc mixed data=Work.TempDataSorted covtest;
	class area breed f_age ageclass no_lambs farm year ID interval ud_size farm;
	model areakm2= area breed year ageclass /solution DDFM= SATTERTHWAITE;
	lsmeans area breed year ageclass /diff;
	random ID /S;
	by interval ud_size;
run;

proc delete data=Work.TempDataSorted;
run;

[Nicolai70]

Hi.

I have used this code to generate a n by n run and obtain 19 times 6 results:

Is it possible to create a matrix output of the type 3 tests only in the form of "interval" by "ud_size" and highlight the significant (on a 0.05 level) effects? I am ony interested in the P values.

Or, should i use another method?

Nicoai

proc sort data=ARTICLE2.TEMP_5_60_SPAT_5_95 out=Work.TempDataSorted;
	by interval ud_size ;
run;

proc mixed data=Work.TempDataSorted covtest;
	class area breed f_age ageclass no_lambs farm year ID interval ud_size farm;
	model areakm2= area breed year ageclass /solution DDFM= SATTERTHWAITE;
	lsmeans area breed year ageclass /diff;
	random ID /S;
	by interval ud_size;
run;

proc delete data=Work.TempDataSorted;
run;

[ballardw]

You might use ODS Output to capture the results of the table(s) of interest and pipe the results into a report procedure that let's you use some of the display options to highlight results.

[Nicolai70]

Do you have any good sources of knowledge on this?