This repository contains jupyter notebooks that are used to perform validation testing of the JWST Calibration Pipeline (hereafter referred to as the calibration pipeline). These notebooks are structured to capture documentation, code, tests, figures, and other outputs that JWST instrument teams use to validate data from the calibration pipeline.

This repository is integrated with the JWST pipeline software repository and the Jenkins automation server. To see most recent build status, go to the STScI Jenkins server.

To see the current suite of validation notebooks, visit our website. Please note that the website is currently only available to internal STScI staff who are on the VPN. Contact Misty Cracraft (@cracraft) or Alicia Canipe (@aliciacanipe) for questions or access issues.

You must be an internal user in order to execute the Validation Notebooks locally, because the test data is only available internally. In order to execute the notebooks or test a new notebook, you can use the following setup:

1. Clone the validation notebooks repository: git clone https://github.com/spacetelescope/jwst_validation_notebooks.git
2. Open a terminal window into the newly created directory
3. Create the Validation Notebooks conda environment: conda env create -f environment.yml
4. Activate the environment: conda activate jwst_validation_notebooks
5. Set up ipykernel for Jupyter: python -m ipykernel install --user --name=jwst_validation_notebooks

You should now be able to run the notebooks with jupyter (type jupyter notebook), or test the creation and running of the test suite by typing python convert.py.

Start jupyter with:

jupyter notebook

This will open a jupyter instance in your web browser, where you can access the notebooks by selecting the jwst_validation_notebooks folder.

From there, you can select the specific testing directory and notebook. jupyter notebooks have an .ipynb extension.

To change the kernel you are using, select the Kernel drop down button in the top left corner of the notebook and hover over "Change Kernel".

From there, you can select the conda environment kernel that includes your JWST pipeline installation. Then, you should be able to execute the notebook. For more information about jupyter notebooks, see the Jupyter Notebook documentation. There is also a handy cheat sheet with shortcuts and commands.

If you would like to generate HTML outputs locally, make sure you are in the jwst_validation_notebooks repository and execute the following commands:

python convert.py

In order to get a full list of run instructions, run

python convert.py --help

There are, however, a few flags that could be especially useful, and a few notes about convert.py which need to be kept in mind.

• In the main notebook directory is a file named exclude_notebooks. This file is currently passed to the --exclude flag if that flag is not set at the command line. As such, it is not currently possible to use the --include flag when running from the main notebook directory.
• To run a subset of notebooks, use some combination of the --notebook-path command-line option to only run the notebooks in a particular directory and the --exclude option to avoid running particular individual notebooks.

Prior to contributing to jwst_validation_notebooks development, please review our style guide. Note that notebook cell outputs must be cleared prior to submission.

Make sure to follow the template outlined in our repository. More information about storing test data is included below.

Before your notebook will be accepted, you must test it in the proper environment described in the Executing Validation Notebooks Locally section above. This will help ensure a smoother delivery of new tests.

This repository operates using the standard fork and pull request github workflow. The following is a bare bones example of this work flow for contributing to the project:

1. Create a fork off of the spacetelescope jwst_validation_notebooks repository.
2. Make a local clone of your fork.
3. Ensure your personal fork is pointing upstream properly.
4. Create a branch on that personal fork.
5. Make your notebook changes and be sure to clear the outputs from the cells.
6. Push that branch to your personal GitHub repository (i.e. origin).
7. On the spacetelescope jwst_validation_notebooks repository, create a pull request that merges the branch into spacetelescope:master.
8. Ensure that the pull request passes the continuous integration check.
9. Assign a reviewer from the team for the pull request (Misty Cracraft @cracraft or Alicia Canipe @aliciacanipe).
10. Iterate with the reviewer over any needed changes until the reviewer accepts and merges your branch.
11. Iterate with the reviewer over copying your test data into either Box or Artifactory.
12. Delete your local copy of your branch.

In order to avoid conflicts between multiple notebooks in the same directory (especially when being run by an automated process), the template notebook contains a cell that sets up a temporary directory and moves the notebook execution into that directory. Even if you don't start your notebook as a copy of the template, you should copy this cell. For development purposes, you may wish to set the use_tempdir variable to False, but when you are ready to submit the notebook in a pull request, please change it to True.

The Jenkins instance is running on a virtual machine inside STScI, so it works best with its CRDS cache set to "/grp/crds/cache", but especially when working over the VPN this location may not work best for you. This cell in the template notebook allows for multiple options as to where to put the CRDS cache directory, so please include it. Again, when submitting the notebook, please set use_local_crds_cache to False.

If you have a notebook that uses updated test data or new test data, follow the steps below to request a data update.

Artifactory should be used for data that is for internal use only.

1. Create a Jira "Task" Issue in the JWST Simulations Jira project requesting to have your data added to Artifactory. Assign the ticket to Misty Cracraft (@cracraft) or Alicia Canipe (@aliciacanipe), and provide more information about the data: simulation information, data location, and pipeline step(s). Once your data has been added to Artifactory, Misty Cracraft (@cracraft) or Alicia Canipe (@aliciacanipe) will resolve the issue and notify you that your data is ready to be used (the full path to the data will be provided by the person who notified you that your data was ingested successfully).

2. Make sure you have the proper OS environmental variable set to access STScI's instance of Artifactory. This can be done via command line or put into a setup file like a .bash_profile file. If you are working in the jwst_validation_notebooks environment, your environment will be set up automatically.

   export TEST_BIGDATA=https://bytesalad.stsci.edu/artifactory/
   

3. Make sure your environment has ci_watson installed. This is done automatically by the jwst_validation_notebooks environment.

   pip install ci_watson
   

4. In your notebook, import the ci_watson package needed.

   from ci_watson.artifactory_helpers import get_bigdata
   

5. Read in each file stored in Artifactory (the full path should have been provided by the person who ingested the data).

   satfile = get_bigdata('jwst_validation_notebooks',
                         'validation_data',
                         'jump',
                         'jump_miri_test',
                         'miri_sat_55k.fits')
   

6. Follow the normal workflow for contributing a notebook once you have confirmed that your notebook is running successfully.

Artifactory is only accessible to internal users on the STScI network. If you would like to contribute a test notebook that uses externally available data, this test data should be stored in our Box folder (jwst_validation_notebooks) instead.

1. Create a Jira "Task" Issue in the JWST Simulations Jira project requesting to have your data added to the Box folder. Assign the ticket to Misty Cracraft (@cracraft) or Alicia Canipe (@aliciacanipe), and provide more information about the data: simulation information, data location, and pipeline step(s). Once your data has been added to Box, Misty Cracraft (@cracraft) or Alicia Canipe (@aliciacanipe) will resolve the issue and notify you that your data is ready to be used (the Box link to the data will be provided by the person who notified you that your data was ingested successfully).
2. Then, in your validation notebook, you will use the following command to import your file from Box (we are using an example file link, you will substitute yours):

from astropy.utils.data import download_file

your_file_box_url ="https://stsci.box.com/shared/static/tpks98b3voqg7r13jt8i6el3yfg9dqoc.fits"
file = download_file(your_file_box_url)

Box assigns a default alphanumerical string as the filename, so you may want to update the filename before processing, or verify that the format is correct. Depending on the data, you can try:

# open file into correct format and write to local disk for processing
with fits.open(file) as hdu:
  hdu.info()
  hdu.writeto(filename)

or use a jwst datamodel:

from jwst.datamodels import RampModel
model = RampModel(file)
model.save(filename)

Users and contributors to the jwst_validation_notebooks repository should adhere to the Code of Conduct. Any issues or violations pertaining to the Code of Conduct should be brought to the attention of a jwst_validation_notebooks team member or to [email protected].

For any questions about the jwst_validation_notebooks project or its software or documentation, please open an Issue.

• Alicia Canipe @aliciacanipe
• Misty Cracraft @cracraft
• Maria Pena-Guerrero @penaguerrero
• Brian York @york_stsci