p*****0
发帖数: 3104 | 1 可以用sas吗？如果可以，用什么procedure？Thanks a lot. |
l*******n
发帖数: 13 | 2 proc freq;
proc genmode;
proc catmod;
....... |
p*****0
发帖数: 3104 | 3 thanks
I only know they can be used to model logistic regression, nice to know that
by the way, which ones are preferred if dealing with large data sets?
【在 l*******n 的大作中提到】
: proc freq;
: proc genmode;
: proc catmod;
: .......
|
l*******n
发帖数: 13 | 4 the procedures deal with count data, the raw dataset should be aggregated
before analysis if it is too large. for example
A B RESPONSE
1 1 0
1 1 0
1 0 1
1 0 0
.....
TO
A B RES0 RES1
1 1 SUM(0) SUM(1)
...
PROC GENMOD IS BEST FOR LOG LINEAR MODEL (multiple responses), can also be
used to fit logit model, allowing predictor variables
PROC FREQ gives simple statistics to represent the association, not for real
-sense modeling |
p*****0
发帖数: 3104 | 5 thanks so much
wonderful!
【在 l*******n 的大作中提到】
: the procedures deal with count data, the raw dataset should be aggregated
: before analysis if it is too large. for example
: A B RESPONSE
: 1 1 0
: 1 1 0
: 1 0 1
: 1 0 0
: .....
: TO
: A B RES0 RES1
|
