0

I am trying to fit a pareto distribution to the following data

x <- 
    c(5857.33154195937, 2352.13410311605, 5868.4139887638, 5084.43835650941, 
      5544.58859069637, 3469.38719024777, 5935.54817351679, 849.551931991455, 
      4804.65544410752, 5707.79165932392, 4073.60262243151, 4828.50927019369, 
      5070.7443738169, 5080.41108830053, 4541.34252127723)

library(fitdistrplus)
library(actuar) # needs r 4.1.0
  
paretoMLE <- fitdist(x, 'pareto', method = 'mle')
  
Error in fitdist(x, "pareto", method = "mle") : 
    the function mle failed to estimate the parameters, 
  with the error code 100

I don't know why this error is happening.

89_Simple
  • 3,393
  • 3
  • 39
  • 94
  • From the fitdist source code: if (mle$convergence>0) stop("the function mle failed to estimate the parameters, with the error code ", mle$convergence, "\n") – Bill O'Brien Jul 20 '21 at 13:56
  • 1
    So the statistics question is why didn't the maximum likelihood estimator converge? Have you tried different x inputs, perhaps a longer vector? – Bill O'Brien Jul 20 '21 at 13:57
  • [with error code 100](https://cran.r-project.org/web/packages/fitdistrplus/vignettes/FAQ.html#the-optimization-algorithm-stops-with-error-code-100.-what-shall-i-do) three suggested steps to getting there. – Chris Jul 20 '21 at 15:26
  • thank you for the pointers. I will have a read – 89_Simple Jul 21 '21 at 11:51
  • I still couldn't figure out why the maximum likelihood estimator did not converge. I added some dummy numbers in my vector increasing it's length but still the same error – 89_Simple Jul 23 '21 at 00:13

0 Answers0