How do you find best probability density distribution for a data set?

Jim21

New Member
#1
How do you find the best probability density distribution that fits a particular data set? Is there any software that will take a data set and run through the various distributions to find the best fit?

I typically work with four distributions: (1) Uniform, (2) Normal, (3) LogNormal and (4) Triangular.

Thanks,

Jim21
 

Dragan

Super Moderator
#2
How do you find the best probability density distribution that fits a particular data set? Is there any software that will take a data set and run through the various distributions to find the best fit?

I typically work with four distributions: (1) Uniform, (2) Normal, (3) LogNormal and (4) Triangular.

Thanks,

Jim21
I think a better - or more flexible - option would be to use a family of distributions with solvable parameters that fit a pdf to a specific set of data.

An example would be making use of Generalized Lambda Distributions (GLDs).

See my first post in this thread:

http://talkstats.com/showthread.php?t=8747&highlight=generalized+lambda
 
#3
hi ,I'm writing a paper about using the application GLD(generalized lambda distributiom) in Insurance,for this paper I've gathered the 39000 data from an insurance company ,i want to use the GLDEX package of R programming ,but i found that it answers for the data that is less than 100! so i want to know that how can I use this test for my data ,or is there any proper test to compare the variouse method of fitting GLD on my data .
thanks