Dear collegs, need help
A problem w. PHREG. are there any advice what does it mean
The last variable always gives no results. DF=0 and no other values.
The code is like
PROC PHREG data=AAA;
MODEL PHDAYS*DIH(1)=AGE SEX PCRE1 PCRE2 PCRE3 PCRE4 PCRE5 PCRE6/RL;
run;
very simple
data like
The SAS System 18:04 Friday, September 18, 2009 105
Cumulative Cumulative
PCRE1 Frequency Percent Frequency Percent
---------------------------------------------------
0 800 83.3 800 83.3
1 160 16.7 960 100.0
Cumulative Cumulative
PCRE2 Frequency Percent Frequency Percent
---------------------------------------------------
0 803 83.6 803 83.6
1 157 16.4 960 100.0
Cumulative Cumulative
PCRE3 Frequency Percent Frequency Percent
---------------------------------------------------
0 807 84.1 807 84.1
1 153 15.9 960 100.0
Cumulative Cumulative
PCRE4 Frequency Percent Frequency Percent
---------------------------------------------------
0 804 83.8 804 83.8
1 156 16.3 960 100.0
Cumulative Cumulative
PCRE5 Frequency Percent Frequency Percent
---------------------------------------------------
0 802 83.5 802 83.5
1 158 16.5 960 100.0
Cumulative Cumulative
PCRE6 Frequency Percent Frequency Percent
---------------------------------------------------
0 784 81.7 784 81.7
1 176 18.3 960 100.0
THE RESULT IS
The PHREG Procedure
Data Set: WORK.AAA
Dependent Variable: PHDAYS RANK FOR VARIABLE HDAYS
Censoring Variable: DIH
Censoring Value(s): 1
Ties Handling: BRESLOW
Summary of the Number of
Event and Censored Values
Percent
Total Event Censored Censored
959 207 752 78.42
Testing Global Null Hypothesis: BETA=0
Without With
Criterion Covariates Covariates Model Chi-Square
-2 LOG L 2745.091 2694.626 50.465 with 7 DF (p=0.0001)
Score . . 52.443 with 7 DF (p=0.0001)
Wald . . 48.693 with 7 DF (p=0.0001)
Analysis of Maximum Likelihood Estimates
Conditional Risk Ratio and
95% Confidence Limits
Parameter Standard Wald Pr > Risk
Variable DF Estimate Error Chi-Square Chi-Square Ratio Lower Upper Label
AGE 1 0.011676 0.00798 2.14021 0.1435 1.012 0.996 1.028 AGE
SEX 1 0.175988 0.15395 1.30686 0.2530 1.192 0.882 1.612 SEX
PCRE1 1 -0.830596 0.23520 12.47090 0.0004 0.436 0.275 0.691
PCRE2 1 -1.311773 0.28412 21.31581 0.0001 0.269 0.154 0.470
PCRE3 1 -0.870943 0.24467 12.67077 0.0004 0.419 0.259 0.676
PCRE4 1 -0.774639 0.23232 11.11825 0.0009 0.461 0.292 0.727
PCRE5 1 -0.039880 0.19120 0.04351 0.8348 0.961 0.661 1.398
PCRE6 0 0 . . . . . .
Is PCRE a categorical variable with 6 levels?
If so, then your model is overspecified and you need only 5 of the levels and the 6th is the reference level.
>Is PCRE a categorical variable with 6 levels?
Yes it is.
But it is separated in to 6 separate vars with vals 0 or 1.
>If so, then your model is overspecified and you need only 5 of the levels and the 6th is the reference level.
If I leave only 5 vars the problem is solved. You are wright.
Thanks.
Consider using the CLASS statement for situations like this; it addresses the overspecification issues for you.
Join us for SAS Innovate April 16-19 at the Aria in Las Vegas. Bring the team and save big with our group pricing for a limited time only.
Pre-conference courses and tutorials are filling up fast and are always a sellout. Register today to reserve your seat.
Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.
Find more tutorials on the SAS Users YouTube channel.