I want to create a 60x3 matrix of integers. Each integer can be in the range 0-40. Each row vector itself in this matrix must have different elements. The vectors itself can be repeated.
What I want:
np.array([[5,6,7],[3,4,10],[20,45,30],[5,6,7]]) #repititions are ok
What I do not want:
np.array([[5,6,7],[3,4,10],[20,45,30],[5,5,20]]) #the 5 is contained twice in the last vector
Using the numpy function: np.random.choice(40, [60, 3], replace=True) yields to cases which I mentioned as "I do not want". Setting replace=False, does not work due to the lack of combinations.
A very inperfomant way so far (but works as wanted):
for i in range(60):
rand[i] = np.random.choice(40, [1, 3], replace=False)
Do you have any recommandations (preferred very perfomant ways)?
