Python#confidence Interval‘札记

import numpy as np
import scipy as sp
import scipy.stats
import numpy

def confidenceinterval(array):
        confidence = 0.95
        a = 1.0*np.array(array)
        m = np.mean(a)
        fc = scipy.stats.sem(a)
        h = fc * sp.stats.t._ppf((1+confidence)/2., n-1)/((n-1)**0.5)
        return  m-h,m+h

n=10
array=numpy.random.uniform(0,1,size=n)
print(confidenceinterval(array))

(0.6110678104984761, 0.7197643000777587)