DTW DynaCLR Monorepo

.gitignore

@@ -66,3 +66,6 @@ slurm*.out
 lightning_logs/
 
 # NOTE: uv.lock is NOT ignored - it should be tracked for reproducibility
+
+# Local-only planning docs (not for upstream)
+applications/dynaclr/docs/DAGs/evaluation_matrix.md

applications/airtable/configs/prepare_config.yml

@@ -0,0 +1,47 @@
+# Dataset preparation pipeline: NFS -> VAST rechunked zarr v3
+# Usage: prepare run <dataset_name> -c prepare_config.yml [--dry-run]
+
+nfs_root: /hpc/projects/intracellular_dashboard/organelle_dynamics
+vast_root: /hpc/projects/organelle_phenotyping/datasets
+workspace_dir: /hpc/mydata/eduardo.hirata/repos/viscy
+
+concatenate:
+  # null = auto-detect raw channels (Phase3D + raw *). Set explicitly to override.
+  channel_names: null
+  chunks_czyx: [1, 16, 256, 256]
+  shards_ratio: [1, 1, 8, 8, 8]
+  output_ome_zarr_version: "0.5"
+  conda_env: biahub
+  # Override biahub's internal SLURM settings (passed via -sb flag)
+  # Set to null to use biahub defaults
+  sbatch_overrides:
+    partition: cpu
+
+qc:
+  channel_names: [Phase3D]
+  NA_det: 1.35
+  lambda_ill: 0.450
+  pixel_size: 0.1494
+  midband_fractions: [0.125, 0.25]
+  device: cuda
+  num_workers: 16
+
+preprocess:
+  channel_names: -1
+  num_workers: 32
+  block_size: 32
+
+# biahub concatenate submits its own SLURM jobs via submitit (no config needed)
+# QC and preprocess run as separate SLURM jobs (no race condition)
+slurm:
+  qc:
+    partition: gpu
+    gres: "gpu:1"
+    cpus_per_task: 16
+    mem_per_cpu: 4G
+    time: "00:30:00"
+  preprocess:
+    partition: cpu
+    cpus_per_task: 32
+    mem_per_cpu: 4G
+    time: "04:00:00"

applications/airtable/scripts/write_experiment_metadata.py

@@ -68,6 +68,9 @@ def register(position_paths: list[Path], dry_run: bool = False, dataset: str | N
         if result.updated:
             db.batch_update(result.updated)
             logger.info("Updated %d existing records", len(result.updated))
+        if result.template_ids_to_delete:
+            db.batch_delete(result.template_ids_to_delete)
+            logger.info("Deleted %d well template records", len(result.template_ids_to_delete))
 
     print(format_register_summary(result, dry_run=dry_run))
 

applications/airtable/src/airtable_utils/database.py

@@ -143,3 +143,18 @@ def batch_create(self, records: list[dict]) -> list[dict]:
             Created records as returned by the Airtable API.
         """
         return self._table.batch_create([r["fields"] for r in records])
+
+    def batch_delete(self, record_ids: list[str]) -> list[dict]:
+        """Batch-delete records by ID.
+
+        Parameters
+        ----------
+        record_ids : list[str]
+            Airtable record IDs to delete.
+
+        Returns
+        -------
+        list[dict]
+            Deletion confirmations from the Airtable API.
+        """
+        return self._table.batch_delete(record_ids)