I have data as below: 5 categories: Delay0, 2, 4, 8 and 16. There are 6 patients in each category. In each category: Patient1 to 4, were measured from day1 to day7; Patient5 was measured from day1 to day6; Patient6 was measured from day1 to day4; I try to use SAS code as follow: Proc mixed; class delay day subject; model value=delay day delay*day / s; repeated day/ type=ar(1); run; When I run the code above there is no error or warning, but I am not so sure it is right. Will Proc mix consider patient5 and patient6 as missing value? Thanks. Message was edited by: Ken_oy ... View more

By using "plot=roc (id=prob)" option, i can create figure1, but this is jpg format graph. I need a png or tiff format for my paper. FIGURE1 I try to use proc gplot to do this, and created two graphs, fig2 and fig3. My questions, 1. how can i merge two graphs in to one (make it the same as fig1). 2. Or any way to create tiff png graphs from proc logistic directly? Thanks!! Figure2 Figure3 ... View more

I try to use "pointlabel" statement to point out ID # of all the outliner, but the problem is it will also list all the ID # of other non-outliner. Is there any conditional statement to control this? Thanks! I want the scatter plot as below, outliner with ID #, non-outliner with a 'star' only. ... View more

Hi there, we can use "call symput" to put data or number under a special word. but how can we clear all these memories. Since I have to put "call symput" in macros and do loops. I want to make sure all the "word"s I will used have been cleaned. Thanks! ... View more