PROC UNIVARIATE: add a quantile-quantile plot for two variables
Ratings are subjective. A thread rated high may not be interesting to the visitor.
On the other hand, a badge for each member designating their SAS expertise level will be useful. Begineer - Intermediate - Advanced - Expert.
I don't recall seeing this poll before, although I may have and simply don't remember. I agree that ratings are subjective, but here is a subjective rating that I think people might find useful: being able to rate responses to any thread.
Often, I post either doesn't get marked as correct, gets marked as correct but actually there is a more correct response, or there are two or more equally correct responses.
Especially when trying to find answers that may have been posted in response to a particular question, I think that many of us would like to be able to quickly narrow down the discussions to just those that others have considered as being otimal responses.
To clarify, you are asking for a scatter plot that shows the quantiles of X1 vs the distributional quantiles AND the quantiles of X2 vs the distributional quantiles? I assume you realize that you can get data values out of the ODS graphics that PROC UNIVARIATE produces, and then use PROC SGPLOT to overlay the Q-Q plots.
No, this request is for a quantile-quantile plot of two variables against each other to compare distributions. This reproduces the example on the NIST web site. 18.104.22.168. Quantile-Quantile Plot
input Run Lab Batch Y X1 X2 X3;
1 1 1 608.781 -1 -1 -1
2 1 2 569.670 -1 -1 -1
3 1 1 689.556 -1 -1 -1
960 8 2 689.942 1 -1 -1
proc sort data=CeramicStrength out=Batch;
by Batch Y;
proc transpose data=Batch out=temp;
proc transpose data=temp out=Batch(rename=(_1=Batch1 _2=Batch2) drop=_Name_);
proc sgplot data=Batch noautolegend;
scatter y=Batch1 x=Batch2;
series y=Batch2 x=Batch2;
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.