topic Help, my K-means clustering is not allowing me to filter my data in Statistical Procedures

Help, my K-means clustering is not allowing me to filter my data

whyamihere — Sun, 30 Mar 2025 00:57:18 GMT

I am trying to filter my data set to cluster only records with the cause of death to be cancer or coronary heart disease. I have tried re-wording the filter several times with the exact same results. Any ideas? TIA

What hints do you have?the filter is appliedthe filter is gone when program ran

proc stdize data=WORK.PRINCOMP_SCORES_HEART_DEAD out=Work._std_ method=range;
	var Prin1 Prin2 Prin3;
run;

proc fastclus data=Work._std_ maxclusters=2 maxiter=20 drift distance 
		out=work.Fastclus_scores_PCA outstat=work.Fastclus_stats_PCA 
		outseed=work.Fastclus_seeds_PCA;
	where "DeathCause in ('Cancer', 'Coronary Heart Disease')";
	var Prin1 Prin2 Prin3;
run;

proc delete data=Work._std_;
run;

Re: Help, my K-means clustering is not allowing me to filter my data

PaigeMiller — Sun, 30 Mar 2025 10:34:54 GMT

where "DeathCause in ('Cancer', 'Coronary Heart Disease')";

Do not put double quotes (or single quotes) around the the WHERE condition. This is also probably why the filter doesn't work in your first screen shot.

It should be

where DeathCause in ('Cancer', 'Coronary Heart Disease');

Re: Help, my K-means clustering is not allowing me to filter my data

whyamihere — Sun, 30 Mar 2025 17:14:36 GMT

Thank you