How to generate a normally distributed variable in Python?-CodePudding

I have a list of 10 values:

variable=[2.1, 5.3, 4.1, 6.7, 2, 6.6, 1.9, 4.51, 4, 7.15]

Its length:

>>> len(variable)
10

Its average:

>>> mean(variable)
4.436

Its standard deviation:

>>> np.std(variable)
1.8987269419271429

From it I want to generate a new_variable having len(new_variable)==100 and normally distributed where mean==4.436 and np.std==1.898.

CodePudding user response：

You can use the random.gauss function:

1 sample:

import random
x = random.gauss(4.436, 1.898)

or 100 samples:

import random
x = [random.gauss(4.436, 1.898) for _ in range(100)]

This is standard library, you don't need to install anything. You may also be interested in the statistics library.

CodePudding user response：

This function achieves exactly what you asked for by transforming a randomly generated distribution.

from statistics import NormalDist, mean, stdev

def get_target_dist(target_mean, target_std, size):
    dist = NormalDist(target_mean, target_std).samples(size)
    dist_mean, dist_std = mean(dist), stdev(dist)
    dist_standard = [(val - dist_mean) / dist_std for val in dist]
    dist_scaled = [val * target_std   target_mean for val in dist_standard]
    return dist_scaled

dist = get_target_dist(4.436, 1.898, 100)

print("len:", len(dist))
print("mean:", mean(dist))
print("std:", stdev(dist))

# len: 100
# mean: 4.436
# std: 1.898