Re: Plotting lognormal distribution

[DimitryASuplatov]

Ok, I`ve figured out how to do the fitting, but now I have the problem
with plotting two graphs at once.

I do:
> pdfNormal = 'my data'
> pdfMean = mean(pdfNormal);
> pdfStd = std(pdfNormal);
> xRange = floor(min(pdfNormal)):0.1:ceil(max(pdfNormal));
> pdfNormalFitted = normpdf(xRange, pdfmean, pdfstd);

Now if I type the following

>h1 = subplot(2,1,1); hist(pdfNormal);
>h2 = subplot(2,1,2); plot(xRange, pdfNormalFitted);
I get two different plots on one sheet. If I run subplot(1,1,1) two
times I get my plots reinitialized so that I see only one graph in the
end.

What I want is - two graphs in one coordinates. Is it possible in
octave?

Thanks!

On Tue, 2008-06-10 at 14:31 +0400, DimitryASuplatov wrote:
> Hello,
> I have 100000 values log-normally distributed. I can calculate its
mean
> and variance to create an "ideal" plot. Though I have some extreme
> values in my sample and I want to see them on the graph too. 
> 
> My question is: how to convert raw data to log-normal distribution
plot
> and draw exactly that plot, not the "ideal" one (by "ideal" plot I
mean
> the one calculated from PDF with specified mean and variance).
> 
> Thanks.