scipy - How to fit a poisson distribution with seaborn? -

i try fit data poisson distribution:

import seaborn sns import scipy.stats stats  sns.distplot(x, kde = false, fit = stats.poisson) 

but error:

attributeerror: 'poisson_gen' object has no attribute 'fit'

other distribution (gamma, etc) de work well.

the poisson distribution (implemented in scipy scipy.stats.poisson) discrete distribution. discrete distributions in scipy not have fit method.

i'm not familiar seaborn.distplot function, appears assume data comes continuous distribution. if case, if scipy.stats.poisson had fit method, not appropriate distribution pass distplot.

---

the question title "how fit poisson distribution seaborn?", sake of completeness, here's 1 way plot of data , fit. seaborn used bar plot, using @mwaskom's suggestion use seaborn.countplot. fitting trivial, because maximum likelihood estimation poisson distribution mean of data.

first, imports:

in [136]: import numpy np  in [137]: scipy.stats import poisson  in [138]: import matplotlib.pyplot plt  in [139]: import seaborn 

generate data work with:

in [140]: x = poisson.rvs(0.4, size=100) 

these values in x:

in [141]: k = np.arange(x.max()+1)  in [142]: k out[142]: array([0, 1, 2, 3]) 

use seaborn.countplot plot data:

in [143]: seaborn.countplot(x, order=k, color='g', alpha=0.5) out[143]: <matplotlib.axes._subplots.axessubplot @ 0x114700490> 

the maximum likelihood estimation of poisson parameter mean of data:

in [144]: mlest = x.mean() 

use poisson.pmf() expected probability, , multiply size of data set expected counts, , plot using matplotlib. bars counts of actual data, , dots expected counts of fitted distribution:

in [145]: plt.plot(k, poisson.pmf(k, mlest)*len(x), 'go', markersize=9) out[145]: [<matplotlib.lines.line2d @ 0x114da74d0>]