Unit tests for makers, container and component modules (part 1/X) #163

guillaumegrolleron · 2025-01-09T08:55:00Z

This PR is the first part of the implementation of the unit tests for :

nectarchain.data.container
nectarchain.makers (ChargesMakers, WaveformsMakers)

It also includes modification of the core containers class to make more robust the I/O methods. Thus the pedestal workflow has been changed to fully use the Necrtachain containers (instead of the pytable module).

The LightNectarcamEventSource has been removed because you are now using the implementation of ctapipe_io_nectarcam

There are also improvement of user scripts for waveforms, charges and photostat computation

ds to find computed calibration quatities

…anted one -update pp and n values to last studies -display wfs -> mean instead of sum -Agg backend is not forced in Pstat calibnration

…ain into dev_test_doc

exluse user_scripts to pytest

improvement split_run method

…nectarcam

pyproject.toml

src/nectarchain/dqm/charge_integration.py

src/nectarchain/makers/calibration/tests/test_pedestal_tool.py

src/nectarchain/user_scripts/ggrolleron/gain_PhotoStat_computation.py

jlenain · 2025-01-15T17:01:19Z

Hi @guillaumegrolleron ,

Many thanks for this PR !

Just one question: did you check that Despeina's GUI and scripts for the TRR are still running fine ? I understand you refactored these scripts to avoid any clash with pytest.

Otherwise, ready to merge or do you want to add things in this PR ?

jlenain

OK... As is, the GUI for the TRR cannot be launched properly. @guillaumegrolleron , could you please either revert your changes in Despina's scripts, or fix it ? At the moment, the different imports are failing. Quickly hacking, but not cleanly, into it, for instance the linearity test fails with:

2025-01-15 21:04:27,329 �[1;32mINFO�[0m [tools_components.LinearityTestTool] (tool.finish): Goodbye
2025-01-15 21:04:27,329 �[1;33mWARNING�[0m [tools_components.LinearityTestTool] (core.finish): Shutting down.
Traceback (most recent call last):
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/linearity.py", line 450, in <module>
    main()
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/linearity.py", line 134, in main
    output = tool.finish()
             ^^^^^^^^^^^^^
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/tools_components.py", line 255, in finish
    dataset = group["ChargeContainer"]
              ~~~~~^^^^^^^^^^^^^^^^^^^
  File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
  File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
  File "/home/jlenain/local/opt/mambaforge/envs/nectar-dev/lib/python3.11/site-packages/h5py/_hl/group.py", line 357, in __getitem__
    oid = h5o.open(self.id, self._e(name), lapl=self._lapl)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
  File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
  File "h5py/h5o.pyx", line 241, in h5py.h5o.open
KeyError: "Unable to synchronously open object (object 'ChargeContainer' doesn't exist)"

I guess this is related to the refactoring from ChargeContainer to ChargeContainers, is that correct ?

jlenain · 2025-01-15T21:31:07Z

If you allow me to commit here, I have a patch ready at least to fix the imports for the TRR GUI test suite. I have put that into a new branch in my fork: https://github.com/jlenain/nectarchain/tree/trr-gui-fix

jlenain · 2025-01-15T22:08:32Z

I guess this is related to the refactoring from ChargeContainer to ChargeContainers, is that correct ?

Hum... not related to that. The same kind of error occurs on the deadtime test:

Traceback (most recent call last):
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/src/deadtime.py", line 424, in <module>
    main()
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/src/deadtime.py", line 135, in main
    output = tool.finish()
             ^^^^^^^^^^^^^
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/src/tools_components.py", line 1246, in finish
    dataset = group["UCTSContainer"]
              ~~~~~^^^^^^^^^^^^^^^^^
  File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
  File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
  File "/home/jlenain/local/opt/mambaforge/envs/nectar-dev/lib/python3.11/site-packages/h5py/_hl/group.py", line 357, in __getitem__
    oid = h5o.open(self.id, self._e(name), lapl=self._lapl)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
  File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
  File "h5py/h5o.pyx", line 241, in h5py.h5o.open
KeyError: "Unable to synchronously open object (object 'UCTSContainer' doesn't exist)"

jlenain · 2025-01-16T10:21:58Z

I guess this is related to the refactoring from ChargeContainer to ChargeContainers, is that correct ?

Hum... not related to that. The same kind of error occurs on the deadtime test:

Traceback (most recent call last):
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/src/deadtime.py", line 424, in <module>
    main()
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/src/deadtime.py", line 135, in main
    output = tool.finish()
             ^^^^^^^^^^^^^
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/src/tools_components.py", line 1246, in finish
    dataset = group["UCTSContainer"]
              ~~~~~^^^^^^^^^^^^^^^^^
  File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
  File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
  File "/home/jlenain/local/opt/mambaforge/envs/nectar-dev/lib/python3.11/site-packages/h5py/_hl/group.py", line 357, in __getitem__
    oid = h5o.open(self.id, self._e(name), lapl=self._lapl)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
  File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
  File "h5py/h5o.pyx", line 241, in h5py.h5o.open
KeyError: "Unable to synchronously open object (object 'UCTSContainer' doesn't exist)"

OK, got it... It is linked to the change from XXXContainer to XXXContainer_0, but caused by what ?

guillaumegrolleron · 2025-01-16T11:45:39Z

Hi @jlenain !
Indeed I forgot to reactive the TRR tests, however the location of this tests in user_scripts is maybe not the best part, what do you think to move this part of the code in a nectarchain module, with associated tests runs by pytest ?
I can do it and fix errors linked to the changes in container I/O methods.

guillaumegrolleron · 2025-01-16T11:47:35Z

I guess this is related to the refactoring from ChargeContainer to ChargeContainers, is that correct ?

Hum... not related to that. The same kind of error occurs on the deadtime test:

Traceback (most recent call last):
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/src/deadtime.py", line 424, in <module>
    main()
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/src/deadtime.py", line 135, in main
    output = tool.finish()
             ^^^^^^^^^^^^^
  File "/home/jlenain/local/src/python/cta-observatory/nectarchain/src/nectarchain/user_scripts/dmousadi/TRR_scripts/src/tools_components.py", line 1246, in finish
    dataset = group["UCTSContainer"]
              ~~~~~^^^^^^^^^^^^^^^^^
  File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
  File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
  File "/home/jlenain/local/opt/mambaforge/envs/nectar-dev/lib/python3.11/site-packages/h5py/_hl/group.py", line 357, in __getitem__
    oid = h5o.open(self.id, self._e(name), lapl=self._lapl)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
  File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
  File "h5py/h5o.pyx", line 241, in h5py.h5o.open
KeyError: "Unable to synchronously open object (object 'UCTSContainer' doesn't exist)"

OK, got it... It is linked to the change from XXXContainer to XXXContainer_0, but caused by what ?

This change is done the make more robust the I/O methods for containers and to have the same behaviour with one or multiple slices.

jlenain · 2025-01-16T11:48:45Z

Hi @guillaumegrolleron !

Concerning the errors linked to the changes in container I/O methods, I am currently working on fixes for it in the branch https://github.com/jlenain/nectarchain/tree/trr-gui-fix. I'll open a PR on top on your PR, if manageable, for that.

jlenain · 2025-01-16T11:49:10Z

Whether the TRR test suite should be kept on user_scripts or not, I have no strong preference about it.

guillaumegrolleron · 2025-01-16T11:51:09Z

Hi @guillaumegrolleron !

Concerning the errors linked to the changes in container I/O methods, I am currently working on fixes for it in the branch https://github.com/jlenain/nectarchain/tree/trr-gui-fix. I'll open a PR on top on your PR, if manageable, for that.

Ok so we first merge this PR and then yours ?

jlenain · 2025-01-16T11:53:39Z

Hi @guillaumegrolleron !
Concerning the errors linked to the changes in container I/O methods, I am currently working on fixes for it in the branch https://github.com/jlenain/nectarchain/tree/trr-gui-fix. I'll open a PR on top on your PR, if manageable, for that.

Ok so we first merge this PR and then yours ?

Yep, we can do that. Green light on your side that I merge this PR ?

guillaumegrolleron · 2025-01-16T11:54:51Z

Yes you can merge !

…R test suite

* Try to fix TRR GUI test suite following PR #163 * Fixed bug in selection of fit function to be used for linearity

guillaume.grolleron added 30 commits June 23, 2024 15:55

add arg to load only wfs and do not extract charges

eb8db7f

cleaner way to treat containers inside makers

2b8b613

new methods to find computed calibration quatities

5f30851

ds to find computed calibration quatities

update to support new version of EvB v6

d8aa001

-bug fix SPE combined : method use to update parameters was not the w…

d183536

…anted one -update pp and n values to last studies -display wfs -> mean instead of sum -Agg backend is not forced in Pstat calibnration

Merge branch 'main' into dev_test_doc

175afa9

add find photostat result method

b46a959

user scripts

75f6b34

Merge branch 'dev_test_doc' of github.com:guillaumegrolleron/nectarch…

04d60f6

…ain into dev_test_doc

add workflow_display to CI

7452e60

change tests name in user_scripts which is not a test for pytest

5f98800

fix import errors in TRR

08b956c

exluse user_scripts to pytest

bugfix : logger

2dfdd82

improvement split_run method

static method decorator were missing

c5d4419

bugfix test containers

404f189

unit test fore nectarchain.makers .core

8c2ac41

bugfix with new version of LIghtNectarCAMEventSource from ctapipe_io_…

413c9b4

…nectarcam

very small code improvement

f023695

Unit test for WaveformsNectarCAMCalibrationTool

37141c2

Unit test for the ChargesNectarCAMCalibrationTool

7308cda

formatting

9e9ee9e

fix test core makers

f11d0c6

formatting

e944206

formatting

02ee8c3

flake8 formatting

d392b4c

refactoring folling pep8

d40c122

folling refactring

65c196d

dqm flake8

24d8ce4

refactoring containers

d571cb3

fix bug formatting

9b9a63c

pytest don't like to have 2 test files with same names

36a7737

jlenain reviewed Jan 9, 2025

View reviewed changes

pyproject.toml Show resolved Hide resolved

pyproject.toml Show resolved Hide resolved

jlenain reviewed Jan 9, 2025

View reviewed changes

src/nectarchain/dqm/charge_integration.py Outdated Show resolved Hide resolved

jlenain reviewed Jan 9, 2025

View reviewed changes

src/nectarchain/makers/calibration/tests/test_pedestal_tool.py Show resolved Hide resolved

jlenain reviewed Jan 9, 2025

View reviewed changes

src/nectarchain/user_scripts/ggrolleron/gain_PhotoStat_computation.py Outdated Show resolved Hide resolved

specific noqa

6dd1e8d

jlenain approved these changes Jan 15, 2025

View reviewed changes

jlenain requested changes Jan 15, 2025

View reviewed changes

jlenain added a commit to jlenain/nectarchain that referenced this pull request Jan 15, 2025

Try to fix TRR GUI test suite following PR cta-observatory#163

aeeb355

jlenain approved these changes Jan 16, 2025

View reviewed changes

jlenain merged commit 5151e86 into cta-observatory:main Jan 16, 2025
11 checks passed

jlenain added a commit to jlenain/nectarchain that referenced this pull request Jan 16, 2025

Following PR cta-observatory#163, modify HDF5 groups to be read by TR…

7ec0b00

…R test suite

jlenain mentioned this pull request Jan 16, 2025

TRR test suite fix #166

Closed

jlenain added a commit to jlenain/nectarchain that referenced this pull request Jan 16, 2025

Try to fix TRR GUI test suite following PR cta-observatory#163

40bb609

This was referenced Jan 16, 2025

TRR test suite fix #167

Merged

Flat field #161

Draft

jlenain added a commit to ArmelleJB/nectarchain that referenced this pull request Jan 16, 2025

Fix conflicts following merge of PR cta-observatory#163

cf1a1df

jlenain added a commit that referenced this pull request Jan 17, 2025

TRR test suite fix (#167)

a7e0226

* Try to fix TRR GUI test suite following PR #163 * Fixed bug in selection of fit function to be used for linearity

guillaumegrolleron mentioned this pull request Jan 17, 2025

Unit test (part 2) #168

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Unit tests for makers, container and component modules (part 1/X) #163

Unit tests for makers, container and component modules (part 1/X) #163

guillaumegrolleron commented Jan 9, 2025

jlenain commented Jan 15, 2025 •

edited

Loading

jlenain left a comment

jlenain commented Jan 15, 2025 •

edited

Loading

jlenain commented Jan 15, 2025

jlenain commented Jan 16, 2025

guillaumegrolleron commented Jan 16, 2025

guillaumegrolleron commented Jan 16, 2025

jlenain commented Jan 16, 2025

jlenain commented Jan 16, 2025

guillaumegrolleron commented Jan 16, 2025

jlenain commented Jan 16, 2025

guillaumegrolleron commented Jan 16, 2025

Unit tests for makers, container and component modules (part 1/X) #163

Unit tests for makers, container and component modules (part 1/X) #163

Conversation

guillaumegrolleron commented Jan 9, 2025

jlenain commented Jan 15, 2025 • edited Loading

jlenain left a comment

Choose a reason for hiding this comment

jlenain commented Jan 15, 2025 • edited Loading

jlenain commented Jan 15, 2025

jlenain commented Jan 16, 2025

guillaumegrolleron commented Jan 16, 2025

guillaumegrolleron commented Jan 16, 2025

jlenain commented Jan 16, 2025

jlenain commented Jan 16, 2025

guillaumegrolleron commented Jan 16, 2025

jlenain commented Jan 16, 2025

guillaumegrolleron commented Jan 16, 2025

jlenain commented Jan 15, 2025 •

edited

Loading

jlenain commented Jan 15, 2025 •

edited

Loading