Is my data fit "extreme value distribution" or "normal distribution"?

Question

I have a large data.frame in R. I would like to double if its distribution fit normal distribution or extreme value distribution better

Here is my simplified data.frame.

x <- data.frame(A=c(1,3,1,5,4,5,5,7,3,2,2,1,1,1,4,9,10))

Could you mind to let me know how to do so? Can I did this analysis with R?

Answer 1

Before considering formal tests, you should try plotting your data. For example,

R> #Plot shown below
R> hist(x$A)
#Backs up the plot
R> shapiro.test(x$A)

Shapiro-Wilk normality test

 data:  x$A 
 W = 0.873, p-value = 0.02452

It's clear from the histogram, that the data doesn't seem to be Normal. I find that it's helpful to get an idea of what's happening before moving onto formal tests.

Does your data fit an extreme valued distribution? That's a bit more tricky. What do you want to do with it?

Answer 2

A good test for normality is the Shapiro-Wilk test which is implemented in R as shapiro.test(x). For general distribution testing there is the Kolmogorov-Smirnov test which is also already implemented in R as ks.test. In addition there is a short introduction in the R manuals about examining data distributions