CosmoStat / CosmoStat/shapepipe

Seed for random in SETools [BUG]

Open
#659 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

bug
Dominant language
Python
Stars
18
Forks
14
Avg merge
8h 40m
Merged PRs (30d)
10

Description

In SETools there is the possibility to select randomly a subpart of a catalogue randomly. When this is done, the random generator is not seeded so we cannot reproduce the results.
This could have an impact on the PSF model (hopefully a small one). If the same exposure is processed twice, and we split the catalogue to have a training and validation set, there will not be the same for both processing. This results in two different PSF model for the same exposure.

A seed should be added in the config file or/and a way to generate a seed based on the exposure name to avoid this behavior.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by locating the SETools code that randomly selects catalogue subsets and the configuration path used when processing an exposure. Reproduce the same exposure twice, then determine whether the seed should come from configuration or the exposure name; done means both runs produce the same catalogue split and PSF model.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.