https://pfspipe.ipmu.jp/jira This file is an XML representation of an issue en-us 8.3.4 803005 13-09-2019 https://pfspipe.ipmu.jp/jira/browse/REDMINE1D-129 1D Redmine <p><em><font color="#505f79"> Created on 2022-11-10 17:58:42 by Vincent Le Brun. % Done: 10</font></em></p> <p>j'ai reçu un nouveau dataset de spectres venus de Fastspec et on dirait qu'ils contiennent des valeurs non numériques qu'on ne gère pas. A confirmer et voir si on veut gérer ce genre de chose<br/> les spectres sont /net/CESAM/amazed/vlebrun/MAMBO/test_MAMBO_for_stacking/output_spectra_sel1DEEP-FORNAX_B30_R10/</p> REDMINE1D-129

[RM-7659] Gestion des pixels non numériques

Task Normal In Progress Unresolved Redmine-Jira Migtation Redmine-Jira Migtation Tue, 13 Jun 2023 05:24:39 +0000 Wed, 22 Nov 2023 19:17:42 +0000 0 2 <p>Comment by Vincent Le Brun on 2022-11-10 17:59:25:<br/> amazed se plante avant de lancer le job slurm avec ce message<br/> /net/CESAM/amazed/venvs/amazed_0.36.1/lib/python3.9/site-packages/pyamazed/amazed.py:38: RuntimeWarning: divide by zero encountered in long_scalars<br/> for _ in range(int(math.ceil(len(iterable)/nb_observations / bunch_size))):<br/> Traceback (most recent call last):<br/> File "/net/CESAM/amazed/venvs/amazed_0.36.1/lib/python3.9/site-packages/pyamazed/amazed.py", line 259, in main<br/> amazed(config)<br/> File "/net/CESAM/amazed/venvs/amazed_0.36.1/lib/python3.9/site-packages/pyamazed/amazed.py", line 99, in amazed<br/> tasks = [(i, f'process_spectra_worker-</p> {i}', spectra)<br/> File "/net/CESAM/amazed/venvs/amazed_0.36.1/lib/python3.9/site-packages/pyamazed/amazed.py", line 99, in &lt;listcomp&gt;<br/> tasks = [(i, f'process_spectra_worker-{i} <p>', spectra)<br/> File "/net/CESAM/amazed/venvs/amazed_0.36.1/lib/python3.9/site-packages/pyamazed/amazed.py", line 38, in bunch<br/> for _ in range(int(math.ceil(len(iterable)/nb_observations / bunch_size))):<br/> OverflowError: cannot convert float infinity to integer<br/> Critical error: cannot convert float infinity to integer</p> <p>Comment by Ali Allaoui on 2022-11-16 16:09:48:<br/> j'ai pas réussi à reproduire ce problème , cf @/net/CESAM/amazed/aallaoui/amazed_workdir/deep_fornax_mambo@</p> <p>Comment by Vincent Le Brun on 2022-11-16 16:38:12:<br/> vizu -p 5002 /net/CESAM/amazed/aallaoui/amazed_workdir/deep_fornax_mambo/output/</p> {'import_dir': ['/net/CESAM/amazed/aallaoui/amazed_workdir/deep_fornax_mambo/output/'], 'reference': None, 'attributes': None, 'title': '', 'filter': None, 'mismatch_path': None, 'config_path': '/net/CESAM/amazed/venvs/amazed_0.36.1/lib/python3.9/site-packages/amazedWebServer/resources/report.json'} <p>Traceback (most recent call last):<br/> File "/net/CESAM/amazed/venvs/amazed_0.36.1/bin/vizu", line 8, in &lt;module&gt;<br/> sys.exit(run())<br/> File "/net/CESAM/amazed/venvs/amazed_0.36.1/lib/python3.9/site-packages/amazedWebServer/app.py", line 297, in run<br/> session_id = mc.import_output_directory(config)<br/> File "/net/CESAM/amazed/venvs/amazed_0.36.1/lib/python3.9/site-packages/amazedWebServer/controller.py", line 41, in import_output_directory<br/> session = Session(import_dir,<br/> File "/net/CESAM/amazed/venvs/amazed_0.36.1/lib/python3.9/site-packages/AmazedOutputAnalyzer/session.py", line 64, in _<em>init</em>_<br/> self.amazed_config<span class="error">&#91;&quot;object_types&quot;&#93;</span><span class="error">&#91;&quot;classification&quot;&#93;</span> = "c"<br/> KeyError: 'object_types'</p> <p>Comment by Vincent Le Brun on 2023-08-29 15:48:37:<br/> ça devrait être géré par les filtres ça ?</p> <p>Comment by Ali Allaoui on 2023-09-19 08:27:31:<br/> Le problème initial survenait à la préparation des bunches, donc soit un bug sur le bunch_size (=0?) ou sur le calcul du nombre d'observations (=0 ?) qui aurait pu entrainer cette exception (on divise par 0 -&gt; inf -&gt; on veut convertir inf en entier)<br/> En tous cas rien à voir avec les pixels non numériques -&gt; on ferme ?</p> Development Rank 0|zzsxlz: