Fitting a Weibull distribution using Scipy

Question

I am trying to recreate maximum likelihood distribution fitting, I can already do this in Matlab and R, but now I want to use scipy. In particular, I would like to estimate

Answer 1

I had the same problem, but found that setting loc=0 in exponweib.fit primed the pump for the optimization. That was all that was needed from @user333700's answer. I couldn't load your data -- your data link points to an image, not data. So I ran a test on my data instead:

import scipy.stats as ss
import matplotlib.pyplot as plt
import numpy as np

N=30
counts, bins = np.histogram(x, bins=N)
bin_width = bins[1]-bins[0]
total_count = float(sum(counts))

f, ax = plt.subplots(1, 1)
f.suptitle(query_uri)

ax.bar(bins[:-1]+bin_width/2., counts, align='center', width=.85*bin_width)
ax.grid('on')
def fit_pdf(x, name='lognorm', color='r'):
    dist = getattr(ss, name)  # params = shape, loc, scale
    # dist = ss.gamma  # 3 params

    params = dist.fit(x, loc=0)  # 1-day lag minimum for shipping
    y = dist.pdf(bins, *params)*total_count*bin_width
    sqerror_sum = np.log(sum(ci*(yi - ci)**2. for (ci, yi) in zip(counts, y)))
    ax.plot(bins, y, color, lw=3, alpha=0.6, label='%s   err=%3.2f' % (name, sqerror_sum))
    return y

colors = ['r-', 'g-', 'r:', 'g:']

for name, color in zip(['exponweib', 't', 'gamma'], colors): # 'lognorm', 'erlang', 'chi2', 'weibull_min', 
    y = fit_pdf(x, name=name, color=color)

ax.legend(loc='best', frameon=False)
plt.show()