Contains fitting for histogram data on normal and lognormal distributions. Though still requiring the fitting toolbox, the parameters are adapted in such a way that Matlab can actually fit the values (e.g. very small x values)!
Functions return mu and sigma parameter, scaling factor (for non-unit distributions) and goodness of fit.
Additionally included are logn2mean and mean2logn to convert between logn parameters mu/sigma and mean/std value (in contrast to lognstat which returns mean and variance)
Nevermind....sorry :-(. I was thinking variance, you are outputting standard deviation. My mistake
I think there might be some math errors in your logn2mean and mean2logn scripts. They do not match up with Matlab's lognstat functions.
I think you need to change line 15 in logn2mean to:
s = exp(2*mu + sigma.^2) .* (exp(sigma.^2) - 1);
And line 14 in mean2logn to:
sigma = sqrt(log(s./m.^2 + 1));
This would then follow the equations laid out my the Mathworks website located at:
Create scripts with code, output, and formatted text in a single executable document.