make_dge yields empty output file

[florensas]

Hello, I am trying to run FICTURE with my own dataset that I pre-processed using /ficture/misc/format_vizgen.py. But when running the make_dge function or run_together I keep getting the same error:

INFO:root:Random` seed 1745334552.4055643
INFO:root:Processing 999938 pixels (1000000 170.68, 18501.14).
/cluster2/florensas/anaconda3/envs/FICTURE/lib/python3.11/site-packages/numpy/_core/fromnumeric.py:3860: RuntimeWarning: Mean of empty slice.
  return _methods._mean(a, axis=axis, dtype=dtype,
/cluster2/florensas/anaconda3/envs/FICTURE/lib/python3.11/site-packages/numpy/_core/_methods.py:145: RuntimeWarning: invalid value encountered in scalar divide
  ret = ret.dtype.type(ret / rcount)

Then the ${path}/hexagon.d_${train_width}.tsv file created is empty and only contain the header. My input data looks like the example examples/data/transcripts.tsv.gz. I tried both using mu scale to 9.24 which is approximately the transformation pixel to um and also manually scaling the coordinates to um and I get the same results.

Did someone face the same issue? Any inputs would be appreciated!

[Yichen-Si]

Hi, apologies for the late reply. I couldn't figure out what caused this error, did you sort your data by one coordinate?
We just made available a new version of FICTURE here, maybe you can give it a try?