Hi all,
I have a data of lognormal distribution (sample size > 1,000,000).
What I want to do is
1) to test if my dataset is a lognormal distribution or not (Histogram shows
a nice normal distribution in log scale but I want to check)
2) two subsets from this dataset have same mean or not (like "t test" of
normal distribution)
What I tried are
1) ad.test{truncgof} of R, which is a "Supremum Class Anderson-Darling test"
but I got error message.
What I had problems are
1) Memory error happen with error message of "cannot allocate vector of size
1.0GB" and I cannot increase the computer spec.
2) I don't know what is "t test" of lognormal.
What I am thinking is
1) Is it OK to convert the data in log scale and do perform t test to say
these dataset has same mean?
How can I solve my problem?
Thank you for any comments.
Hyunchul
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
