I need to fill a length n 1D numpy array with random numbers from the ranges [1,2,...,n], [2,3,...,n], [3,4,...,n], ..., [n-1,n] and [n] respectively. I am looking for a vectorize solution to this problem. Thanks very much in advance.
1
-
1Why does it need to be efficient?wildplasser– wildplasser2021-09-26 17:59:15 +00:00Commented Sep 26, 2021 at 17:59
Add a comment
|
1 Answer
You could use numpy.random.randint, for this:
import numpy as np
n = 10
res = np.random.randint(np.arange(0, n), n)
print(res)
Output
[3 3 2 6 6 7 6 8 9 9]
From the documentation:
Generate a 1 by 3 array with 3 different lower bounds
np.random.randint([1, 5, 7], 10) array([9, 8, 7]) # random
The more up-to-date alternative is to use integers:
import numpy as np
n = 10
rng = np.random.default_rng()
res = rng.integers(np.arange(0, n), n)
print(res)
Note: The examples above start from 0, choose the intervals that suit best your problem.
