CLI Reference¶

SCoPe provides a set of command-line tools installed via the scope-ml package. Below is a categorized reference of all available commands.

Core commands¶

Command	Entry point	Description
`scope-initialize`	`scope.__init__:initialize`	Initialize required directories and copy necessary files to run SCoPe
`scope-develop`	`scope._instantiate:develop`	Set up a development installation of SCoPe
`scope-lint`	`scope.scope_class:Scope.lint`	Run linting checks on the codebase
`scope-doc`	`scope._instantiate:doc`	Generate imagery and build the documentation
`scope-train`	`scope._instantiate:train`	Train a binary classifier (DNN or XGBoost) for a given classification tag
`scope-test`	`scope._instantiate:test`	Run the full test suite (requires Kowalski access)
`scope-test-limited`	`scope._instantiate:test_limited`	Run a limited test suite (no Kowalski access required)

Command	Entry point	Description
`create-training-script`	`scope._instantiate:create_training_script`	Generate a shell script containing multiple `scope-train` calls
`assemble-training-stats`	`scope._instantiate:assemble_training_stats`	Assemble and summarize training statistics across classifiers
`train-algorithm-slurm`	`tools.train_algorithm_slurm:main`	Generate SLURM scripts to train all classifiers in parallel on HPC
`train-algorithm-job-submission`	`tools.train_algorithm_job_submission:main`	Submit SLURM training jobs to the cluster

Command	Entry point	Description
`generate-features`	`tools.generate_features:main`	Generate features for ZTF light curves
`generate-features-slurm`	`tools.generate_features_slurm:main`	Generate SLURM scripts for parallelized feature generation
`generate-features-job-submission`	`tools.generate_features_job_submission:main`	Submit SLURM feature generation jobs to the cluster
`check-quads-for-sources`	`tools.generate_features_slurm:check_quads_for_sources`	Check which quadrants contain sources of interest

Command	Entry point	Description
`run-inference`	`tools.inference:main`	Run trained classifiers on a field of generated features
`create-inference-script`	`scope._instantiate:create_inference_script`	Generate a shell script for running inference across all classifiers
`run-inference-slurm`	`tools.run_inference_slurm:main`	Generate SLURM scripts for parallelized inference
`run-inference-job-submission`	`tools.run_inference_job_submission:main`	Submit SLURM inference jobs to the cluster
`combine-preds`	`tools.combine_preds:main`	Combine DNN and XGB prediction files for a field
`combine-preds-slurm`	`tools.combine_preds_slurm:main`	Generate SLURM scripts for parallelized prediction combining

Command	Entry point	Description
`scope-download-classification`	`tools.scope_download_classification:main`	Download classifications from Fritz/SkyPortal
`scope-upload-classification`	`tools.scope_upload_classification:main`	Upload classifications to Fritz/SkyPortal
`scope-manage-annotation`	`tools.scope_manage_annotation:main`	Manage annotations on Fritz/SkyPortal
`post-taxonomy`	`tools.taxonomy:main`	Post the SCoPe taxonomy to Fritz/SkyPortal
`select-fritz-sample`	`scope._instantiate:select_fritz_sample`	Select a sample of sources from Fritz for labeling or analysis
`get-quad-ids`	`tools.get_quad_ids:main`	Retrieve quadrant IDs for ZTF fields

Command	Entry point	Description
`get-rubin-ids`	`tools.get_rubin_ids:main`	Discover Rubin object IDs via cone search or read from CSV
`generate-features-rubin`	`tools.generate_features_rubin:main`	Generate SCoPe features for a set of Rubin sources
`prepare-rubin-chunks`	(see README)	Scan local parquet files and split eligible objects into chunk CSVs
`generate-features-rubin-slurm`	(see README)	Generate a SLURM array job script from chunk files
`combine-rubin-features`	(see README)	Merge per-chunk parquet outputs into a single combined file

Command	Entry point	Description
`run-scope-local`	`tools.run_scope_local:main`	Run the SCoPe classification pipeline locally on a set of sources
`analyze-logs`	`tools.analyze_logs:main`	Analyze log files from SCoPe pipeline runs