# [R] log(0) problem in max likelihood estimation

Bill Simpson wsi at gcal.ac.uk
Tue Jan 9 17:32:46 CET 2001

```This practical problem in maximum likelihood estimation must be
encountered quite a bit. What do you do when a data point has a
probability that comes out in numerical evaluation to zero? In calculating
the log likelihood you then have a log(0) problem.

Here is a simple example (probit) which illustrates the problem:

x<-c(1,2,3,4,100)
ntrials<-100
yes<-round(ntrials*pnorm((x-3)/1)) #points fall on normal CDF mean=3, sd=1
no<-ntrials-yes
pyes<-yes/ntrials
py<-function(b,x) pnorm((x-b)/b)
loglike<-function(b) -sum( yes*log(py(b,x)) + no*log(1-py(b,x)) )
out<-nlm(loglike,p=c(3,1),hessian=TRUE)

In this example the right-most point gives a p(yes) of 1; 1-1=0; log(0)
gives "NA/Inf replaced by maximum positive value"

Please tell me how to deal with this problem.  Thanks very much for any
help.

Bill Simpson

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

```