[R] stablefit can fit the parameters of a truncated normal distribution?

[drbn]

Thanks John ,

I did look at this function but seems that only works with censored (not
truncated) normal distributions.

I know that a straightforward to do it is with fitdistr from MASS package:



library(MASS) #fitdistr
library(msm) #dtnorm, rtnorm

#Generate truncated normal sample
data<-rtnorm(1000,mean=108, sd=5, lower=105, upper= 136)

#Trucated normal distribution
dtnorm0 <- function(X, mean, sd, log = FALSE) {dtnorm(X, mean, sd, 105, 136,
log)}

#Fit data to normal
tres <- fitdistr(data,"normal")

#Fit data to truncated normal
nres <- fitdistr(data, dtnorm0, start=list(mean=mean(data), sd=sd(data)))

#results
tres
nres



When 'data' came from a generated truncated normal distribution there is no
problem. My problem is that I want to fit a real sample collected over 10
years. For each year,I only have the biased mean of the sample (like 'tres')
and I want to estimate the mean of the population (like 'nres'). When I try
to do it, I get correct estimates of the underlying normal distribution in
some years but in others there is a message "optimization failed".

I know that the underlying distribution is normal but some years my sample
data does not look like this.

I don't know how to get an estimation of the parameters in these
"optimization failed" years and this is the reason to seek for some
alternatives to 'fitdistr', but of course I can be wrong.

Thanks

David


John C Frain wrote:
> 
> This idea is very wrong,  Have a look at the function cnormal1 in the
> VGAM package
> 
> John frain
> 
> 2008/10/15 drbn <drbn at yahoo.com>:
>>
>> I'm using stableFit from the package fBasics to estimate the parameters
>> of a
>> truncated normal distribution (I'm interested in the parameters of the
>> underlying normal distribution). It is correct to generalize this
>> truncated
>> normal distribution as a stable distribution ?
>>
>> Thanks
>>
>> David
>> --
>> View this message in context:
>> http://www.nabble.com/stablefit-can-fit-the-parameters-of-a-truncated-normal-distribution--tp20002335p20002335.html
>> Sent from the R help mailing list archive at Nabble.com.
>>
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide
>> http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>
> 
> 
> 
> -- 
> John C Frain
> Trinity College Dublin
> Dublin 2
> Ireland
> www.tcd.ie/Economics/staff/frainj/home.html
> mailto:frainj at tcd.ie
> mailto:frainj at gmail.com
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
> 
> 

-- 
View this message in context: http://www.nabble.com/stablefit-can-fit-the-parameters-of-a-truncated-normal-distribution--tp20002335p20028459.html
Sent from the R help mailing list archive at Nabble.com.