Fibrosis models

[AOkenov]

Different way of modeling fibrosis (removing points, connections, cells)

[hansdierckx]

I would like to make a very thin lesion by removing edges from the domain, such that all vertices of the grid still have a readout. Is this possible?
(For my current goal, removing edges transverse to straight line will do).

Thanks.

[AOkenov]

@hansdierckx
We will add this as fibrosis model in future releases. For now you can use following approach:

set barrier indices
x_min = n // 4
x_max = 3 * n // 4
y = n//4
x = np.arange(x_min, x_max)
y0 = np.ones_like(x) * y
y1 = np.ones_like(x) * (y + 1)

initialize model
model.initialize()
remove corresponding connection between (x, y0) and (x, y1)
0 - is weights of the node (i-1, j)
1 - is weights of the node (i, j-1)
2 - is weights of central node (i, j)
3 - is weights of node (i, j+1)
4 - is weights of node (i+1, j)

model.weights[x, y0, 2] += model.weights[x, y0, 3]
model.weights[x, y0, 3] = 0
model.weights[x, y1, 2] += model.weights[x, y1, 1]
model.weights[x, y1, 1] = 0

run the model without initialization
model.run(initialize=False)