Closed carljohnsen closed 5 months ago
Not a problem anymore, but this problem was encountered during #25 as well: it happened when the value range was smaller than the histogram size (vmax-vmin < voxel_bins
), in which case the discretization ensured that certain bins could never be filled.
If we run
hist.sum(axis=0)
, the curve looks like spiky, with the spikes growing alongside the magnitude of the curve.It only happens if we remove ~half of the scan on the y-axis (which we do due to the samples being cut in half).