You can generate the array normally as I explained in my comment:
r_tot = []
for r_i in however_many_rs_there_are: # each r_i contains an array of 6 values
r_tot.append(r_i)
You can then convert r_tot into a numpy
array like so:
import numpy
np_array = numpy.array(r_tot)
Here's a very simple proof of concept:
>>> import random, numpy
>>> r_tot = []
>>> for i in range(0,random.randint(1,20)): # append an arbitrary number of arrays
r_i = [1,2,3,4,5,6] # all of size six
r_tot.append(r_i) # to r_tot
>>> np_array = numpy.array(r_tot) # then convert to numpy array!
>>> np_array # did it work?
array([[1, 2, 3, 4, 5, 6],
[1, 2, 3, 4, 5, 6],
[1, 2, 3, 4, 5, 6],
[1, 2, 3, 4, 5, 6],
[1, 2, 3, 4, 5, 6],
[1, 2, 3, 4, 5, 6], # yeaaaah
[1, 2, 3, 4, 5, 6],
[1, 2, 3, 4, 5, 6],
[1, 2, 3, 4, 5, 6],
[1, 2, 3, 4, 5, 6],
[1, 2, 3, 4, 5, 6],
[1, 2, 3, 4, 5, 6],
[1, 2, 3, 4, 5, 6]])