Cylinders not randomly distributed in x-y plane from generators.cylinders

Due to some property of how random fiber locations are selected, fibers are more likely to cross near the center of the x-y plane versus point closer to the edge. This code makes a structure, then calculates the porosity at each x-y location.

import porespy as ps
import matplotlib.pyplot as plt
import numpy as np

im = ps.generators.cylinders([400,400,1000],5, 7000,0,90,None)
mean = np.mean(im,2)
plt.imshow(mean,cmap='plasma')
plt.show()
print(ps.metrics.porosity(im))

Results:

A more dramatic example for a 1000x1000x1000 fiber image with a porosity of 0.5.

[jgostick]

I am going to merge this change into dev since you’re ok with it. I will close this issue now.

[jgostick]

@1minus1 I implemented something close to your quick fix in PR #245. The starting points are generated over a much larger region now. I basically increase the region by the requested cylinder length on all sides. Anything more than that would be pointless since the the generated cylinder would not reach that actual domain. I did this without actually creating a larger image, which is good. The down side, as you say is that more tries are required to get the correct number of cylinders. The function was still decently fast on my laptop though, but I didn’t compare times. Let me know if this new version works better for you. We can still discuss the porosity thing separately, but at least you’ll have a valid function to build on!