topic Re: Class Model Information missing for PROC LOGISTIC in Statistical Procedures

Class Model Information missing for PROC LOGISTIC

lady8506 — Thu, 19 Jul 2018 14:16:09 GMT

I have been using a subset of my data to practice logistic regression and when the results are output, the "Class Level Information" table is missing. I'm not sure why this is. Can someone take a look at my code and my data and tell me what might be causing the problem?

Also, would it be better to use the CLASS statement in PROC LOGISTIC, or to dummy code everything myself instead? At this point, I'm not sure what the difference is.

Here is my code:

TITLE "Logistic Regression: Reconstruction or not";
PROC LOGISTIC DATA = merge_temp7 descending;
CLASS HistologyGroup TumorStageGroup;
MODEL BreastConservingTherapy = Age1 had_chemo; 
RUN;

HistologyGroup can have the following values: 2, 3, 4 Refers to the type of tumor they have

TumorStageGroup can have the following values: 0, 1, 2, 3, 4 Refers to the stage of the tumor

BreastConservingTherapy can be either 1 for 'yes" or 0 for 'no' and refers to the type of surgery a person had

had_chemo is the same as BreastConservingTherapy. Did the patient receive chemo or not?

Age1 is a person's age, so it's continuous.

Here is some sample data from those variables:

SAS Output

HistologyGroup TumorStageGroup BreastConservingTherapy Age1 had_chemo

2	3	0	66	1
4	2	0	64	1
4	3	0	46	1
2	2	1	60	1
4	2	1	60	1
4	1	0	84	0
4	1	1	59	0
4	1	0	46	1
4	1	1	56	0
3	0	1	68	0
4	1	1	51	0
3	0	0	51	0
4	1	.	40	.
4	1	0	42	1
4	2	0	64	0
3	.	0	35	0
4	3	0	61	1
4	2	0	49	1
4	2	0	49	.
4	1	0	40	0
4	2	1	71	0
4	1	1	54	1
4	1	1	77	0
4	3	0	47	1
4	3	0	75	1
4	2	1	53	0
3	0	1	52	0
3	0	1	66	0
4	1	1	72	.

Re: Class Model Information missing for PROC LOGISTIC

PaigeMiller — Thu, 19 Jul 2018 14:24:21 GMT

You want to use

MODEL BreastConservingTherapy = Age1 had_chemo HistologyGroup TumorStageGroup;

Also, do not try to code dummy variables to use instead of using the CLASS statement. Creating your own dummy variables is a total waste of time and an opportunity to make mistakes.

Re: Class Model Information missing for PROC LOGISTIC

Ksharp — Thu, 19 Jul 2018 14:26:36 GMT

You didn't include the class variables in your model.

PROC LOGISTIC DATA = merge_temp7 descending;
CLASS HistologyGroup TumorStageGroup;
MODEL BreastConservingTherapy = Age1 had_chemo    HistologyGroup TumorStageGroup;

Re: Class Model Information missing for PROC LOGISTIC

lady8506 — Thu, 19 Jul 2018 14:45:20 GMT

Thank you for your help. What an easy fix. I feel silly now!

2	3	0	66	1
4	2	0	64	1
4	3	0	46	1
2	2	1	60	1
4	2	1	60	1
4	1	0	84	0
4	1	1	59	0
4	1	0	46	1
4	1	1	56	0
3	0	1	68	0
4	1	1	51	0
3	0	0	51	0
4	1	.	40	.
4	1	0	42	1
4	2	0	64	0
3	.	0	35	0
4	3	0	61	1
4	2	0	49	1
4	2	0	49	.
4	1	0	40	0
4	2	1	71	0
4	1	1	54	1
4	1	1	77	0
4	3	0	47	1
4	3	0	75	1
4	2	1	53	0
3	0	1	52	0
3	0	1	66	0
4	1	1	72	.

2	3	0	66	1
4	2	0	64	1
4	3	0	46	1
2	2	1	60	1
4	2	1	60	1
4	1	0	84	0
4	1	1	59	0
4	1	0	46	1
4	1	1	56	0
3	0	1	68	0
4	1	1	51	0
3	0	0	51	0
4	1	.	40	.
4	1	0	42	1
4	2	0	64	0
3	.	0	35	0
4	3	0	61	1
4	2	0	49	1
4	2	0	49	.
4	1	0	40	0
4	2	1	71	0
4	1	1	54	1
4	1	1	77	0
4	3	0	47	1
4	3	0	75	1
4	2	1	53	0
3	0	1	52	0
3	0	1	66	0
4	1	1	72	.

2	3	0	66	1
4	2	0	64	1
4	3	0	46	1
2	2	1	60	1
4	2	1	60	1
4	1	0	84	0
4	1	1	59	0
4	1	0	46	1
4	1	1	56	0
3	0	1	68	0
4	1	1	51	0
3	0	0	51	0
4	1	.	40	.
4	1	0	42	1
4	2	0	64	0
3	.	0	35	0
4	3	0	61	1
4	2	0	49	1
4	2	0	49	.
4	1	0	40	0
4	2	1	71	0
4	1	1	54	1
4	1	1	77	0
4	3	0	47	1
4	3	0	75	1
4	2	1	53	0
3	0	1	52	0
3	0	1	66	0
4	1	1	72	.