Solved
New Contributor
Posts: 2

PCA using spearman rank correlation

Hello,

I am trying to find some correlations between 11 variables (10 continuous variables (with 2 not normally distributed) and 1 categorial variable (1, 2, 3 or 4)). For that, I would like to start with a descriptive analysis using PCA. Regarding the kind of variables (and distributions) I have, I would like to perform a PCA using a matrix of Spearman's correlation. I have seen that we can do it with R software but I'm using SAS and I don't find any information about it, how to do it...

Do you have any idea of how I can do it (in which statement, syntax).

Thank you very much.

Sophie

Accepted Solutions
Solution
‎10-07-2015 10:49 AM
Esteemed Advisor
Posts: 5,535

Re: PCA using spearman rank correlation

Try, for example

``````proc corr data=sashelp.cars nomiss outs=carsSpear;
var invoice--length;
run;

proc princomp data=carsSpear n=2 plots=pattern;
var invoice--length;
run;``````
PG

All Replies
Super User
Posts: 23,754

Re: PCA using spearman rank correlation

Look into PROC PRINCOMP and/or PROC FACTOR.
Solution
‎10-07-2015 10:49 AM
Esteemed Advisor
Posts: 5,535

Re: PCA using spearman rank correlation

Try, for example

``````proc corr data=sashelp.cars nomiss outs=carsSpear;
var invoice--length;
run;

proc princomp data=carsSpear n=2 plots=pattern;
var invoice--length;
run;``````
PG
New Contributor
Posts: 2

Re: PCA using spearman rank correlation

YES!!! It's working!!

Thank you very much!

Sophie

🔒 This topic is solved and locked.

Need further help from the community? Please ask a new question.

Discussion stats
• 3 replies
• 706 views
• 1 like
• 3 in conversation