Python testing if my data follows a lognormal distribution

Question

Hi I have a distribution of results that is positively skewed so I want to test if it is a good fit to a log-normal distribution or a Gumbell distribution.

I have us

Answer 1

You can simply use scipy.stats.lognorm.fit to fit your data to a lognormal distribution. This will give you a tuple

(0.60845558877160033, 0.27409944344131409, 1.8037732130179509)

which represents shape, location, and scale respectively. If you want the more common parameters of mu and sigma, you can obtain them like so

shape, location, scale = scipy.stats.lognorm.fit(listofdata)
mu, sigma = np.log(scale), shape

You can use the scipy.stats.gumbel_l.fit function similarily for that respective distribution, which will return the location and scale.

As far as how to test goodness of fit for each distribution, one possibility would be to use the Kolmogorov-Smirnov test with scipy.stats.kstest.