Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Setup integration tests pipeline #70

Closed
14 tasks done
HealthyPear opened this issue Oct 23, 2020 · 2 comments · Fixed by #137
Closed
14 tasks done

Setup integration tests pipeline #70

HealthyPear opened this issue Oct 23, 2020 · 2 comments · Fixed by #137
Labels
good first issue Good for newcomers help wanted Extra attention is needed maintenance summary A summary of issues related to the same subject
Projects
Milestone

Comments

@HealthyPear
Copy link
Member

HealthyPear commented Oct 23, 2020

These are NOT unit-tests.
These are NOT benchmarks.

These are tests checking that all protopipe scripts (aka the "steps" or "tools" of the pipeline) end correctly with proper output.

Current status

Operations to perform

Each integration test should be contained in a template function that performs the following checks,

  • the script doesn't crash
  • the output file exists and it is not empty (provided that the input settings are correct)
  • the output file has the correct format (probably this will be checked indirectly through the ctapipe and pyirf libraries)

Requirements

  • an analysis configuration (the set of protopipe's config files)
    • DL1/DL2
    • Energy model
    • Classification model
    • DL3
  • a server from which to store the simulated test dataset(s)
  • setup the CI that steers these operations

Dataset(s)

We should have 2 datasets (CTA-N + CTA-S both baseline arrays) per production.

Each dataset could have the same composition as a normal analysis, so

  • 10% gammas for energy training
  • 10% gammas + 60% protons for classification training
  • 80% gammas + 40% protons + 100% electrons for performance estimation

UPDATE: Actually it is not necessary, we can just use the same proportions between files and at the same time keeping the current pipeline workflow

They could be produced directly using simtel and not by cutting an already existing simtel file.
This to avoid situations in which a small test file doesn't contain enough triggered events.
So we could simulate e.g. 100 showers (=10% statistics) with E > 10TeV within a scatter radius of 150 meters.

UPDATE: Documentation has been updated accordingly (at least for Prod3b test files)

@HealthyPear HealthyPear added this to To do in Maintenance via automation Oct 23, 2020
@HealthyPear HealthyPear added the help wanted Extra attention is needed label Feb 6, 2021
@HealthyPear HealthyPear added the summary A summary of issues related to the same subject label Feb 19, 2021
@HealthyPear HealthyPear moved this from To do to Summary issues in Maintenance Feb 19, 2021
@HealthyPear
Copy link
Member Author

Opened cta-observatory/ctapipe#1651 to test the download of testdata from the cc-in2p3 dataserver.

@HealthyPear
Copy link
Member Author

HealthyPear commented Mar 10, 2021

after PR #111

so we would be able to

  • test on diffuse data only
  • test already on prod5

@HealthyPear HealthyPear added this to the v0.5.0 milestone Apr 14, 2021
@HealthyPear HealthyPear changed the title Integration tests Setup integration tests pipeline May 13, 2021
@HealthyPear HealthyPear linked a pull request May 13, 2021 that will close this issue
Maintenance automation moved this from Summary issues to Done May 14, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
good first issue Good for newcomers help wanted Extra attention is needed maintenance summary A summary of issues related to the same subject
Projects
Maintenance
  
Done
Development

Successfully merging a pull request may close this issue.

1 participant