[R] Question about predict.multinom()

Giovanni Petris GPetris at uark.edu
Tue May 11 22:31:53 CEST 2004


Hello,

This is the fitted model:

> fit
Call:
multinom(formula = resp ~ pred$cls + pred$smoke)

Coefficients:
             (Intercept)  pred$cls2   pred$cls3   pred$cls4  pred$cls5 pred$smoke2 pred$smoke3
Proteinuria    -1.140520  0.1616644  0.05554898 -0.01584927 0.02574805  -0.4057245  -0.2898425
Hypertension   -2.691215 -0.3699690 -0.22582107  0.01615898 0.26318005   0.1239051   0.2413282
Both           -2.285950 -0.4473108 -0.16212932 -0.10477158 0.01272335  -0.4852405  -0.6290152

Residual Deviance: 22809.76 
AIC: 22851.76 

... and these, to my understanding, should be the estimated category
probabilities for a subject in the first `cls' category and 3rd
`smoke' category:

> predict(fit, newdata=data.frame(cls=factor("1",levels=levels(pred$cls)),
+ smoke=factor("3",levels=levels(pred$smoke))), type="probs")
     Neither  Proteinuria Hypertension         Both 
   0.6715336    0.7394394    0.7247787    0.6722327 

Why I am not getting probabilities? Could somebody tell me what I am
missing? 

Thank you in advance,
Giovanni

-- 

 __________________________________________________
[                                                  ]
[ Giovanni Petris                 GPetris at uark.edu ]
[ Department of Mathematical Sciences              ]
[ University of Arkansas - Fayetteville, AR 72701  ]
[ Ph: (479) 575-6324, 575-8630 (fax)               ]
[ http://definetti.uark.edu/~gpetris/              ]
[__________________________________________________]




More information about the R-help mailing list