zkhan zkhan - 5 months ago 44
R Question

Fitting a normal distribution in R

I'm using the following code to fit the normal distribution. The link for the dataset for "b" (too large to post directly) is :

link for b


tazur <-read.csv("b", header= TRUE, sep=",")

enter image description here

After plotting the histogram, it seems a normal distribution should fit well.

f1n <- fitdistr(claims,"normal")

#Length Class Mode
#estimate 2 -none- numeric
#sd 2 -none- numeric
#vcov 4 -none- numeric
#n 1 -none- numeric
#loglik 1 -none- numeric


Error in xy.coords(x, y, xlabel, ylabel, log) :

'x' is a list, but does not have components 'x' and 'y'

I get the above error when I try to plot the fitted distribution, and even the summary statistics are off for f1n.

Would appreciate any help.


Looks like you are making confusion between MASS::fitdistr and fitdistrplus::fitdist.

  • MASS::fitdistr returns object of class "fitdistr", and there is no plot method for this. So you need to extract estimated parameters and plot the estimated density curve yourself.
  • I don't know why you load package fitdistrplus, because your function call clearly shows you are using MASS. Anyway, fitdistrplus has function fitdist which returns object of class "fitdist". There is plot method for this class, but it won't work for "fitdistr" returned by MASS.

I will show you how to work with both packages.

## reproducible example
set.seed(0); x <- rnorm(500)

Using MASS::fitdistr

No plot method is available, so do it ourselves.

fit <- fitdistr(x, "normal")
# [1] "fitdistr"

para <- fit$estimate
#         mean            sd 
#-0.0002000485  0.9886248515 

hist(x, prob = TRUE)
curve(dnorm(x, para[1], para[2]), col = 2, add = TRUE)

enter image description here

Using fitdistrplus::fitdist

FIT <- fitdist(x, "norm")    ## note: it is "norm" not "normal"
# [1] "fitdist"

plot(FIT)    ## use method `plot.fitdist`

enter image description here