similar to: Lognormal with 0/1 "tag" [theoretical]

Hi Again, I think I've solved my problem, but please tell me if you think I'm wrong, or you can see a better way! A plot of the integrand showed a very sharp peak, so I was running into the integrand "feature" mentioned in the note. I resolved it by limiting the range of integration as shown here: -------------------------------------------------- function (x, meanlog = 0,

Hi! I have following data which is left truncated say at 10. I am trying to estimate the parameters of the Truncated Lognormal distribution to this data as given below. (I have referred to R code appearing in an earlier post - http://finzi.psych.upenn.edu/Rhelp10/2008-October/176136.html) library(MASS) x <- c(600.62,153.05,70.26,530.42,3440.29,97.45,174.51,168.47, 116.63,36.51, 219.77,

Hi again, Adding further information to my own query, this function gets to the core of the problem, which I think lies in the behaviour of 'integrate'. ------------------------------------- function (x, meanlog = 0, sdlog = 1, ...) { require(stats) integrand <- function(t, x, meanlog, sdlog) dpois(x,t)*dlnorm(t, meanlog, sdlog) mapply(function(x, meanlog, sdlog, ...) #

Hi, I am trying to learn lognormal mixture models with EM. I was wondering how does one compute the log likelihood. The current implementation I have is as follows, which perform really bad in learning the mixture models. __BEGIN__ # compute probably density of lognormal. dens <- function(lambda, theta, k){ temp<-NULL meanl=theta[1:k] sdl=theta[(k+1):(2*k)]

Hi, just for the record, although I don't think it's relevant (!) ------------------------------------- > sessionInfo() R version 2.6.0 (2007-10-03) i386-pc-mingw32 locale: LC_COLLATE=English_United Kingdom.1252;LC_CTYPE=English_United Kingdom.1252;LC_MONETARY=English_United Kingdom.1252;LC_NUMERIC=C;LC_TIME=English_United Kingdom.1252 attached base packages: [1] stats4 splines

Dear List, I have read that a lognormal mixture model having a pdf of the form f(x)=w1*f1(x)+(1-w1)*f2(x) fits most data sets quite well, where f1 and f2 are lognormal distributions. Any pointers on how to create a function that would produce the 5 parameters of f(x) would be greatly appreciated. > version _ platform i386-pc-mingw32 arch i386 os

I am trying to fit a lognormal distribution to a set of data and test its goodness of fit with regard to predicted values. I managed to get so far: > y <- c(2,6,2,3,6,7,6,10,11,6,12,9,15,11,15,8,9,12,6,5) > library(MASS) > fitdistr(y,"lognormal",start=list(meanlog=0.1,sdlog=0.1)) meanlog sdlog 1.94810515 0.57091032 (0.12765945) (0.09034437) But I would

The density of a lognormal should be 0 for negative arguments, but > dlnorm(-1) [1] NaN Warning message: NaNs produced in: dlnorm(x, meanlog, sdlog, log) A simple fix is to change dlnorm's definition to: function (x, meanlog = 0, sdlog = 1, log = FALSE) .Internal(dlnorm(x*(x>0), meanlog, sdlog, log)) It might be faster to put the same sort of adjustment into the internal code, but

I am trying to plot a lognormal density curve on top of an existing histogram. Can anybody suggest a simple way to do this? Even if someone could just explain how to plot a regular normal density curve on top of an existing histogram, it would be a big help. Also, is there some way to search through the R-help archives other than simple browsing? Thank you so much. Your help and time is greatly

Hi , I am trying to get some estimator based on lognormal distribution when we have left,interval, and right censored data. Since, there is now avalible pakage in R can help me in this, I had to write my own code using Newton Raphson method which requires first and second derivative of log likelihood but my problem after runing the code is the estimators were too high. with this email ,I provide

Dear R-users I've got the next problem: I've got this *function*: fitcond=function(x,densfun,pcorte,start,...){ myfn <- function(parm,x,pcorte,...) -sum(log(dens(parm,x,pcorte,...))) Call <- match.call(expand.dots = TRUE) if (missing(start)) start <- NULL dots <- names(list(...)) dots <- dots[!is.element(dots, c("upper",

Hello, I tried to fit a lognormal distribution by using optim. But sadly the output seems to be incorrect. Who can tell me where the "bug" is? test = rlnorm(100,5,3) logL = function(parm, x,...) -sum(log(dlnorm(x,parm,...))) start = list(meanlog=5, sdlog=3) optim(start,logL,x=test)$par Carsten. [[alternative HTML version deleted]]

Hello, I've the following setting: (1) Data from a source without truncation (x) (2) Data from an other source with left-truncation at threshold u (xu) I have to fit a model on these these two sources, thereby I assume that both are "drawn" from the same distribution (eg lognormal). In a MLE I would sum the densities and maximize. The R-Function could be:

Hello. I am not certain even how to search the archives for this particular question, so if there is an obvious answer, please smack me with a large halibut and send me to the URLs. I have been experimenting with fitting curves by using both maximum likelihood and maximum spacing estimation techniques. Originally, I have been writing distribution-specific functions in 'R' which work

Hi, Does anyone know what the mean value of a lognormal distribution in base-2 is? I am simulating stochastic population growth and if I were working in base-e, I would do:lambda <- 1.1 #multiplicative growth rates <- 0.6 #stochasticity (std. dev)lognormal <- rlnorm(100000, log(lambda) - (s^2)/2, s)## or lognormal <- exp( rnorm( 100000, log(lambda) - (s^2)/2,

Dear all, I want to compare some different models for a dataset by QQ plots and AIC. I get the following AICs: - linear model: 19759.66 - GAMLSS model: 18702.7 - linear model with lognormal response: -7862.182 The QQ plots show that the lognormal model fits better than the linear model, but still much worse than the GAMLSS. So, in my opinion, the AIC of the lognormal model should be between the

The background: I'm trying to fit a Poisson-lognormal distrbutuion to some data. This is a way of modelling species abundances: N ~ Pois(lam) log(lam) ~ N(mu, sigma2) The number of individuals are Poisson distributed with an abundance drawn from a log-normal distrbution. To fit this to data, I need to integrate out lam. In principle, I can do it this way: PLN1 <- function(lam, Count,

Hi, I'm playing around with gamlss and don't entirely understand the sigma result from an attempted lognormal fit. In the example below, I've created lognormal data with mu=10 and sigma=2. When I try a gamlss fit, I get an estimated mu=9.947 and sigma=0.69 The mu estimate seems in the ballpark, but sigma is very low. I get similar results on repeated trials and with Normal and

Hi, I am trying to get the maximum likelihood estimator for lognormal distribution with censored data;when we have left, interval and right censord. I built my code in R, by writing the deriving of log likelihood function and using newton raphson method but my estimators were too high " overestimation", where the values exceed the 1000 in some runing of my code. is there any one can

Dear R users I use simulated data to evaluate a model by sampling the parameters in my model from lognormal distributions. I would like these (lognormal distributed) parameters to be correlated, that is, I would like to have pairwise samples of 2 parameters with a given correlation coefficient. I have seen that a covariance matrix can be fixed when generating random variables from a