a colleague of mine (in CC) and myself are implementing a spatially structured network in NEST (using current NEST master). We are wondering how to generate a population of neurons that are distributed on a regular grid plus a random jitter. This means that we want to end up with a spatial distribution as can be seen in Figure 8 in the tutorial for creating spatially structured networks. Below that figure one finds a code snippet; however, we believe that this does not correspond to the distribution of Figure 8.

In the documentation of NEST 2.20.1 we did find a code snippet that produces a jittered grid:

import numpy as np
# grid with jitter
jit = 0.03
xs = np.arange(-0.5,.501,0.1)
poss = [[x,y] for y in xs for x in xs]
poss = [[p[0]+np.random.uniform(-jit,jit),p[1]+np.random.uniform(-jit,jit)] for p in poss]
layer_dict_ex = {"positions": poss,
        "extent" : [1.1,1.1],
        "elements" : "iaf_psc_alpha"}

As is apparent, this code does not make use of NEST internal random number generation but rather produces a list of positions that is then passed to a layer.

Thus our question is: is it possible to create a jittered grid natively in NEST master?

Best regards,

Jasper and Anno