Avec SAS-GENMOD:
Voici ce que contient le fichier bino:
1 1 17 35
1 2 14 20
1 3 8 25
2 1 12 30
2 2 22 40
2 3 15 50
3 1 24 30
3 2 41 60
3 3 11 20
Programme:
data bino;
infile "bino";
input l c y n;
proc genmod data=bino;
class l c;
model y/n=l c l*c/ dist = binomial
link = cll
type1;
run;
Sortie:
The GENMOD Procedure
Model Information
Description Value
Data Set WORK.BINO
Distribution BINOMIAL
Link Function CLL
Dependent Variable Y
Dependent Variable N
Observations Used 9
Number Of Events 164
Number Of Trials 310
Class Level Information
Class Levels Values
L 3 1 2 3
C 3 1 2 3
Criteria For Assessing Goodness Of Fit
Criterion DF Value Value/DF
Deviance 0 -0.0000 .
Scaled Deviance 0 -0.0000 .
Pearson Chi-Square 0 0.0000 .
Scaled Pearson X2 0 0.0000 .
Log Likelihood . -196.6287 .
Analysis Of Parameter Estimates
Parameter DF Estimate Std Err ChiSquare Pr>Chi
INTERCEPT 1 -0.2250 0.3096 0.5283 0.4673
L 1 1 -0.7278 0.4716 2.3816 0.1228
L 2 1 -0.8059 0.4040 3.9793 0.0461
L 3 0 0.0000 0.0000 . .
C 1 1 0.7009 0.3838 3.3347 0.0678
C 2 1 0.3647 0.3508 1.0809 0.2985
C 3 0 0.0000 0.0000 . .
L*C 1 1 1 -0.1561 0.5787 0.0728 0.7874
L*C 1 2 1 0.7737 0.5745 1.8136 0.1781
L*C 1 3 0 0.0000 0.0000 . .
L*C 2 1 1 -0.3417 0.5476 0.3894 0.5326
L*C 2 2 1 0.4412 0.4882 0.8168 0.3661
L*C 2 3 0 0.0000 0.0000 . .
L*C 3 1 0 0.0000 0.0000 . .
L*C 3 2 0 0.0000 0.0000 . .
L*C 3 3 0 0.0000 0.0000 . .
SCALE 0 1.0000 0.0000 . .
NOTE: The scale parameter was held fixed.
LR Statistics For Type 1 Analysis
Source Deviance DF ChiSquare Pr>Chi
INTERCEPT 35.4480 0 . .
L 15.9369 2 19.5112 0.0001
C 5.3727 2 10.5642 0.0051
L*C -0.0000 4 5.3727 0.2512
_______________________________________________________________
Programme:
proc genmod data=bino;
class l c;
make 'obstats' out=predi;
model y/n=l c/ dist = binomial
link = cll
obstats
covb
corrb
waldci;
run;
data new;
set predi;
proc plot data=new;
plot yvar1*reschi yvar1*pred ;
run;
Sortie:
Criteria For Assessing Goodness Of Fit
Criterion DF Value Value/DF
Deviance 4 5.3727 1.3432
Scaled Deviance 4 5.3727 1.3432
Pearson Chi-Square 4 5.2527 1.3132
Scaled Pearson X2 4 5.2527 1.3132
Log Likelihood . -199.3151 .
Analysis Of Parameter Estimates
Parameter DF Estimate Std Err ChiSquare Pr>Chi
INTERCEPT 1 -0.3579 0.2129 2.8268 0.0927
L 1 1 -0.4555 0.2140 4.5322 0.0333
L 2 1 -0.6645 0.1946 11.6658 0.0006
L 3 0 0.0000 0.0000 . .
C 1 1 0.5488 0.2266 5.8659 0.0154
C 2 1 0.6604 0.2178 9.1973 0.0024
C 3 0 0.0000 0.0000 . .
SCALE 0 1.0000 0.0000 . .
NOTE: The scale parameter was held fixed.
Estimated Covariance Matrix
Parameter
Number PRM1 PRM2 PRM3 PRM5 PRM6
PRM1 0.04531 -0.02020 -0.02171 -0.03135 -0.03681
PRM2 -0.02020 0.04578 0.01722 -0.002222 0.009396
PRM3 -0.02171 0.01722 0.03785 0.004212 0.009330
PRM5 -0.03135 -0.002222 0.004212 0.05135 0.03053
PRM6 -0.03681 0.009396 0.009330 0.03053 0.04743
Estimated Correlation Matrix
Parameter
Number PRM1 PRM2 PRM3 PRM5 PRM6
PRM1 1.0000 -0.4434 -0.5242 -0.6498 -0.7940
PRM2 -0.4434 1.0000 0.4136 -0.0458 0.2017
PRM3 -0.5242 0.4136 1.0000 0.0956 0.2202
PRM5 -0.6498 -0.0458 0.0956 1.0000 0.6187
PRM6 -0.7940 0.2017 0.2202 0.6187 1.0000
Normal Confidence Intervals For Parameters
Two-Sided Confidence Coefficient: 0.9500
Parameter Confidence Limits
PRM1 Lower -0.7751
PRM1 Upper 0.0593
PRM2 Lower -0.8749
PRM2 Upper -0.0361
PRM3 Lower -1.0458
PRM3 Upper -0.2832
PRM5 Lower 0.1047
PRM5 Upper 0.9930
PRM6 Lower 0.2336
PRM6 Upper 1.0873
Observation Statistics
Y N Pred Xbeta Std HessWgt Lower
17 35 0.5358 -0.2646 0.1869 18.7025 0.4127
14 20 0.5761 -0.1530 0.2081 9.4789 0.4349
8 25 0.3581 -0.8134 0.2252 9.0499 0.2481
12 30 0.4635 -0.4736 0.1919 14.1808 0.3479
22 40 0.5016 -0.3619 0.1795 18.4437 0.3873
15 50 0.3021 -1.0224 0.1994 14.9665 0.2160
24 30 0.7019 0.1909 0.1843 16.1985 0.5698
41 60 0.7416 0.3025 0.1383 41.6576 0.6437
11 20 0.5030 -0.3579 0.2129 9.2558 0.3691
Observation Statistics
Upper Resraw Reschi Resdev
0.6694 -1.7544 -0.5946 -0.5937
0.7248 2.4788 1.1216 1.1431
0.4981 -0.9529 -0.3975 -0.4009
0.5963 -1.9064 -0.6980 -0.7011
0.6284 1.9366 0.6124 0.6130
0.4124 -0.1069 -0.0329 -0.0329
0.8240 2.9426 1.1745 1.2205
0.8305 -3.4969 -1.0313 -1.0088
0.6539 0.9402 0.4205 0.4209
Plot of YVAR1*RESCHI. Legend: A = 1 obs, B = 2 obs, etc.
40 + A
|
|
|
|
| A A
Y 20 +
| A A
| A A
| A
| A
|
0 +
--+----------+----------+----------+----------+----------+----------+-
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
Reschi
Plot of YVAR1*PRED. Legend: A = 1 obs, B = 2 obs, etc.
40 + A
|
|
|
|
| A A
Y 20 +
| A A
| A A
| A
| A
|
0 +
--+------------+------------+------------+------------+------------+--
0.3 0.4 0.5 0.6 0.7 0.8
Pred