Galaxy / Dust Masking

[axf295]

allow for using a mask map before map preprocessing, etc.

[Copilot]

Pull request overview

Adds support for applying an external “dusty galaxy” mask map as a configurable preprocessing step prior to other map preprocessing operations.

Changes:

• Introduces a new GalaxyMask map preprocessor that loads/applies a galaxy mask map.
• Extends mask_dustgal to accept an optional logger and emit structured logs.
• Updates preprocessor configuration types to include galaxy_mask, and adjusts the historical lightcurve SLURM wrapper timestamp parsing/header generation.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

File	Description
sotrplib/maps/preprocessor.py	Adds GalaxyMask preprocessor that loads a mask map via pixell.enmap and combines it with existing masks.
sotrplib/maps/masks.py	Updates mask_dustgal signature to accept log and adds logging around mask extraction.
sotrplib/config/preprocessors.py	Adds GalaxyMaskConfig and includes it in the preprocessor config union.
scripts/historical_lightcurve_extractor/slurm_wrapper_depth1_lightcurves.py	Adjusts timestamp parsing and changes SLURM header generation (removing soconda module loading parameter).

Comments suppressed due to low confidence (1)

scripts/historical_lightcurve_extractor/slurm_wrapper_depth1_lightcurves.py:172

• generate_slurm_header() no longer loads the soconda module, but the script still exposes a --soconda-version CLI option (and get_soconda_module() helper) that are now unused. Either reintroduce the module load behavior (using the CLI arg) or remove the unused CLI option/helper to avoid misleading users into thinking their --soconda-version input has an effect.

def generate_slurm_header(jobname, groupname, cpu_per_task, run_dir, slurm_out_dir):
    slurm_header = f"""#!/bin/bash
#SBATCH --job-name={jobname}            # create a short name for your job
#SBATCH -A {groupname}                    # group name, by default simonsobs
#SBATCH --nodes=1                # node count
#SBATCH --ntasks=1               # total number of tasks across all nodes
#SBATCH --cpus-per-task={cpu_per_task}       # cpu-cores per task (>1 if multi-threaded tasks)
#SBATCH --mem-per-cpu=4G         # memory per cpu-core (4G is default)
#SBATCH --time=04:59:00          # total run time limit (HH:MM:SS)
#SBATCH --output={slurm_out_dir}%x.out

export SRUN_CPUS_PER_TASK=$SLURM_CPUS_PER_TASK
cd {run_dir}
source .venv/bin/activate

"""

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

GalaxyMaskConfig.mask_path defaults to None, but to_preprocessor() always constructs GalaxyMask(mask_path=self.mask_path). With a missing mask_path, the preprocessor will fail at runtime when it tries to read the mask file. Consider making mask_path required (no default) or validate in to_preprocessor() and raise a clear configuration error when it's not provided.

Suggested change

	) -> MapPreprocessor:
	) -> MapPreprocessor:
	if self.mask_path is None:
	raise ValueError(
	"GalaxyMaskConfig.mask_path must be set to a valid file path "
	"before creating a GalaxyMask preprocessor."
	)

[Copilot]

GalaxyMask.preprocess will call enmap.read_map(str(self.mask_path)) when mask_map is None, but mask_path can be None (default). This will try to read a file named "None" and fail with a confusing error. Add an explicit validation/error path when both mask_map and mask_path are unset (e.g., raise ValueError with a clear message, or log and return the input map unchanged).

Suggested change

	if self.mask_map is None:
	if self.mask_map is None:
	if self.mask_path is None:
	raise ValueError(
	"GalaxyMask.preprocess requires either 'mask_map' or 'mask_path' to be set."
	)