Open m03077yhtnt opened 7 months ago
Hi @m03077yhtnt
Can you try again after reinstalling numpy
pip install --upgrade numpy --force-reinstall
Best, Seppe
Hi @m03077yhtnt
Can you try again after reinstalling numpy
pip install --upgrade numpy --force-reinstall
Best, Seppe
I have the same error. After upgrading numpy there are more errors related to package incompatibility. ---------------------------------------------------------------------------
WorkerCrashedError Traceback (most recent call last)
Cell In[27], line 6
2 macs_path = "macs2"
4 os.makedirs(os.path.join(out_dir, "consensus_peak_calling/MACS"), exist_ok = True)
----> 6 narrow_peak_dict = peak_calling(
7 macs_path = macs_path,
8 bed_paths = bed_paths,
9 outdir = os.path.join(os.path.join(out_dir, "consensus_peak_calling/MACS")),
10 genome_size = 'hs',
11 n_cpu = 8,
12 input_format = 'BEDPE',
13 shift = 73,
14 ext_size = 146,
15 keep_dup = 'all',
16 q_value = 0.05,
17 _temp_dir = temp_dir
18 )
File ~/.conda/envs/scenicplus3.11/lib/python3.11/site-packages/pycisTopic/pseudobulk_peak_calling.py:286, in peak_calling(macs_path, bed_paths, outdir, genome_size, n_cpu, input_format, shift, ext_size, keep_dup, q_value, nolambda, skip_empty_peaks, **kwargs)
284 except Exception as e:
285 ray.shutdown()
--> 286 raise(e)
287 ray.shutdown()
288 else:
File ~/.conda/envs/scenicplus3.11/lib/python3.11/site-packages/pycisTopic/pseudobulk_peak_calling.py:264, in peak_calling(macs_path, bed_paths, outdir, genome_size, n_cpu, input_format, shift, ext_size, keep_dup, q_value, nolambda, skip_empty_peaks, **kwargs)
262 ray.init(num_cpus=n_cpu, **kwargs)
263 try:
--> 264 narrow_peaks = ray.get(
265 [
266 macs_call_peak_ray.remote(
267 macs_path,
268 bed_paths[name],
269 name,
270 outdir,
271 genome_size,
272 input_format,
273 shift,
274 ext_size,
275 keep_dup,
276 q_value,
277 nolambda,
278 skip_empty_peaks
279
280 )
281 for name in list(bed_paths.keys())
282 ]
283 )
284 except Exception as e:
285 ray.shutdown()
File ~/.conda/envs/scenicplus3.11/lib/python3.11/site-packages/ray/_private/auto_init_hook.py:22, in wrap_auto_init.<locals>.auto_init_wrapper(*args, **kwargs)
19 @wraps(fn)
20 def auto_init_wrapper(*args, **kwargs):
21 auto_init_ray()
---> 22 return fn(*args, **kwargs)
File ~/.conda/envs/scenicplus3.11/lib/python3.11/site-packages/ray/_private/client_mode_hook.py:103, in client_mode_hook.<locals>.wrapper(*args, **kwargs)
101 if func.__name__ != "init" or is_client_mode_enabled_by_default:
102 return getattr(ray, func.__name__)(*args, **kwargs)
--> 103 return func(*args, **kwargs)
File ~/.conda/envs/scenicplus3.11/lib/python3.11/site-packages/ray/_private/worker.py:2626, in get(object_refs, timeout)
2624 raise value.as_instanceof_cause()
2625 else:
-> 2626 raise value
2628 if is_individual_id:
2629 values = values[0]
WorkerCrashedError: The worker died unexpectedly while executing this task. Check python-core-worker-*.log files for more information.
Hi @cjiang310437
Is it feasible to try to run this code with one core?
Best,
Seppe
Hi @cjiang310437
Is it feasible to try to run this code with one core?
Best,
Seppe
Hi @SeppeDeWinter, Thanks for replying. I tried upgrading numpy and one core but both did not work. It looks like it's not the issue of numpy or job parallels. Here is the situation: I have been running scenicplus on our HPC clusters. The code always worked fine. But recently our HPC clusters machines are upgraded from CentOS7 to RHEL9. And this code started to getting the following errors for 'undefined_symbol: log_finite' using the RHEL9 machines. The error occured both on peak_calling(macs2) and purturbation(velocyto) steps. Can I get any idea for resolving the issue?
The error got from peakcalling using scenicplus version downloaded from main branch:
The error got from plot_perturbation_effect_in_embedding using scenicplus version downloaded from old branch:
Any suggestions would be appreciated. Thank you!
Best, Cheng
Hi @cjiang310437
Did you try to reinstall all packages after the OS change? It might be that some packages need to be recompiled to the new os.
All the best,
Seppe
Describe the bug When I attempt to run "peak_calling", I encounter an error, which I believe is related to numpy. I installed scenicplus using Python 3.11 following the instructions on the website: https://github.com/aertslab/scenicplus/tree/development. Based on my understanding, I am using "MACS2-2.2.7.1-py3.8-linux-x86_64". Could the error be caused by the difference in Python versions? I would appreciate any advice you can provide. Thank you so much for your help.
To Reproduce narrow_peaks_dict = peak_calling(macs_path, bed_paths, os.path.join(work_dir, 'scATAC/consensus_peak_calling/MACS/'), genome_size='hs', n_cpu=2, input_format='BEDPE', shift=73, ext_size=146, keep_dup = 'all', q_value = 0.05)
Version (please complete the following information):
Error output