Optimize/Optimize() 非线性 function 用于序数逻辑回归 Function 在 R

Question

I want to maximize this nonlinear function, but I cannot figure out what I am doing incorrectly. Here is the code:

max.fun<-function(y0,y1,y2,pi0,pi1,pi2){y0*log(pi0)+y1*log(pi1)+y2*log(pi2)} 

th = matrix(c(-1,-3,1),nrow=1,ncol = 3)
y0 = matrix(c(26,104),nrow=1,ncol = 2)
y1 = matrix(c(33,72),nrow=1,ncol = 2)
y2 = matrix(c(22,31),nrow=1,ncol = 2)
race = matrix(c(1,0),nrow=1,ncol = 2)

alpha1<-th[1]
alpha2<-th[2]
beta<-th[3]

ex1<-exp(alpha1+beta*race)
ex2<-exp(alpha2+beta*race)

ee1<-ex1/(1+ex1)
ee2<-ex2/(1+ex2)

pi0<-1/(1+ex1)
pi1<-ee1-ee2
pi2<-ee2


optimize(max.fun,
         c(0,1),
         y0=y0,
         y1=y1,
         y2=y2,
         pi0=pi0,
         pi1=pi1,
         pi2=pi2,
         tol = 0.0001)

When I run the optimize function, I get this error message:

Error in f(arg, ...) : unused argument (arg)

Any help is greatly appreciated.

Answer 1

What are you trying to optimize? If x then it needs to be on LHS & other values must be provided- such as if you want to maximize y0:

max.fun<-function(y0,y1,y2,pi0,pi1,pi2){y0*log(pi0)+y1*log(pi1)+y2*log(pi2)} 

th = matrix(c(-1,-3,1),nrow=1,ncol = 3)
y0 = matrix(c(26,104),nrow=1,ncol = 2)
y1 = matrix(c(33,72),nrow=1,ncol = 2)
y2 = matrix(c(22,31),nrow=1,ncol = 2)
race = matrix(c(1,0),nrow=1,ncol = 2)

alpha1<-th[1]
alpha2<-th[2]
beta<-th[3]

ex1<-exp(alpha1+beta*race)
ex2<-exp(alpha2+beta*race)

ee1<-ex1/(1+ex1)
ee2<-ex2/(1+ex2)

pi0<-1/(1+ex1)
pi1<-ee1-ee2
pi2<-ee2

# Using purely illustrative values
y0m <- optimize(max.fun,
         c(0,1),
         y1=2,
         y2=3,
         pi0=4,
         pi1=5,
         pi2=6,
         tol = 0.0001,
         maximum = T)
y0m

> y0m
$maximum
[1] 0.9999339

$objective
[1] 9.980357

If this/any other answer provided to you is useful to you please accept the answer /vote+ !