Routine Weekly Merge (#1055)

* Update config parameters for run_1.

* Fix shape sampling.

* Create bounding boxes for DES.

* Remove file datum generation file.

* Exploring DES outputs.

* Clean up notebook.

* Update inference notebook.

* Modify prior to use SVA1 sources.

* Modify prior to correctly calculate radii in pixels.

* Change default image size to 2560.

* Modify mean sources to match DES.

* Modify default config.

* Modify data gen to write catalogs on the fly.

* Script for basic cluster counting.

* Script for mapping SVA1 regions. Flake8 throws import error.

* Modify prior to use each source redshift instead of single cluster redshift.

Author: kapnadak

case_studies/galaxy_clustering/conf/config.yaml

@@ -5,10 +5,10 @@ defaults:
     - override hydra/job_logging: stdout
 
 data_gen:
-    data_dir: /nfs/turbo/lsa-regier/scratch/kapnadak/new_data
-    image_size: 1280
-    tile_size: 128
-    nfiles: 5000
+    data_dir: /data/scratch/kapnadak/data-08-08/
+    image_size: 2560
+    tile_size: 256
+    nfiles: 2000
     n_catalogs_per_file: 500
     bands: ["g", "r", "i", "z"]
     min_flux_for_loss: 0
@@ -37,7 +37,7 @@ cached_simulator:
     batch_size: 2
     splits: 0:60/60:90/90:100  # train/val/test splits as percent ranges
     num_workers: 8
-    cached_data_path: /nfs/turbo/lsa-regier/scratch/kapnadak/new_data/file_data
+    cached_data_path: /data/scratch/kapnadak/new_data/file_data
     train_transforms: []
 
 train:
@@ -81,7 +81,7 @@ encoder:
 
 predict:
     cached_dataset:
-        _target_: case_studies.galaxy_clustering.cached_dataset.CachedDESModule
+        _target_: case_studies.galaxy_clustering.inference.cached_dataset.CachedDESModule
         cached_data_path: /nfs/turbo/lsa-regier/scratch/gapatron/desdr-server.ncsa.illinois.edu/despublic/dr2_tiles
         tiles_per_img: 64
         batch_size: 2
@@ -100,6 +100,6 @@ predict:
                 output_dir: "/data/scratch/des/dr2_detection_output/run_1"
                 write_interval: "batch"
     encoder: ${encoder}
-    weight_save_path: /nfs/turbo/lsa-regier/scratch/gapatron/best_encoder.ckpt
+    weight_save_path: /data/scratch/des/best_encoder.ckpt
     device: "cuda:0"
     output_save_path: "/data/scratch/des/dr2_detection_output/run_0"

case_studies/galaxy_clustering/data_generation/SVA_map.py

@@ -0,0 +1,33 @@
+# flake8: noqa
+import pickle
+
+import numpy as np
+import pandas as pd
+from astropy.table import Table
+
+SVA_PATH = "/data/scratch/des/sva1_gold_r1.0_catalog.fits"
+BOUNDING_BOX_PATH = "/data/scratch/des/bounding_coordinates.pickle"
+
+
+def main():
+    sva_catalog = Table.read(SVA_PATH).to_pandas()
+    bounding_boxes = pd.read_pickle(BOUNDING_BOX_PATH)
+    des_sva_intersection = []
+    output_filename = "/data/scratch/des/sva_map.pickle"
+
+    for k, v in bounding_boxes.items():
+        ra_min, ra_max, dec_min, dec_max = v["RA_min"], v["RA_max"], v["DEC_min"], v["DEC_max"]
+        ra_intersection = np.logical_and((ra_min < sva_catalog["RA"]), (sva_catalog["RA"] < ra_max))
+        dec_intersection = np.logical_and(
+            (dec_min < sva_catalog["DEC"]), (sva_catalog["DEC"] < dec_max)
+        )
+        full_intersection = np.logical_and(ra_intersection, dec_intersection)
+        if full_intersection.any():
+            des_sva_intersection.append(k)
+
+    with open(output_filename, "wb") as handle:
+        pickle.dump(des_sva_intersection, handle, protocol=pickle.HIGHEST_PROTOCOL)
+
+
+if __name__ == "__main__":
+    main()

case_studies/galaxy_clustering/data_generation/coordinates.py

@@ -0,0 +1,35 @@
+import os
+import pickle
+from pathlib import Path
+
+import pandas as pd
+from astropy.io import fits
+
+
+def main():
+    des_dir = Path(
+        "/nfs/turbo/lsa-regier/scratch/gapatron/desdr-server.ncsa.illinois.edu/despublic/dr2_tiles/"
+    )
+    des_subdirs = [d for d in os.listdir(des_dir) if d.startswith("DES")]
+    bounding_coordinates = {}
+    output_filename = "/data/scratch/des/bounding_coordinates.pickle"
+    for des_subdir in des_subdirs:
+        catalog_path = des_dir / Path(des_subdir) / Path(f"{des_subdir}_dr2_main.fits")
+        catalog_data = fits.getdata(catalog_path)
+        source_df = pd.DataFrame(catalog_data)
+        ra_min, ra_max = source_df["RA"].min(), source_df["RA"].max()
+        dec_min, dec_max = source_df["DEC"].min(), source_df["DEC"].max()
+        bounding_box = {
+            "RA_min": ra_min,
+            "RA_max": ra_max,
+            "DEC_min": dec_min,
+            "DEC_max": dec_max,
+        }
+        bounding_coordinates[des_subdir] = bounding_box
+
+    with open(output_filename, "wb") as handle:
+        pickle.dump(bounding_coordinates, handle, protocol=pickle.HIGHEST_PROTOCOL)
+
+
+if __name__ == "__main__":
+    main()

case_studies/galaxy_clustering/data_generation/data_gen.py

@@ -48,16 +48,14 @@ def catalog_gen(cfg):
     cluster_prior = ClusterPrior(image_size=image_size)
     background_prior = BackgroundPrior(image_size=image_size)
 
-    combined_catalogs = []
-    for _ in range(nfiles):
+    for i in range(nfiles):
         background_catalog = background_prior.sample_background()
         if np.random.uniform() < 0.5:
             cluster_catalog = cluster_prior.sample_cluster()
-            combined_catalogs.append(pd.concat([cluster_catalog, background_catalog]))
+            catalog = pd.concat([cluster_catalog, background_catalog])
         else:
-            combined_catalogs.append(background_catalog)
-
-    for i, catalog in enumerate(combined_catalogs):
+            catalog = background_catalog
+        print(f"Writing catalog {i} ...")
         file_name = f"{catalogs_path}/{file_prefix}_{i:03}.dat"
         catalog_table = Table.from_pandas(catalog)
         astro_ascii.write(catalog_table, file_name, format="no_header", overwrite=True)

case_studies/galaxy_clustering/data_generation/file_datum_generation.py

@@ -0,0 +1,59 @@
+import os
+import pickle
+
+import torch
+
+DES_DIR = (
+    "/nfs/turbo/lsa-regier/scratch/gapatron/desdr-server.ncsa.illinois.edu/despublic/dr2_tiles"
+)
+DES_BANDS = ("g", "r", "i", "z")
+DES_SUBDIRS = (d for d in os.listdir(DES_DIR) if d.startswith("DES"))
+OUTPUT_DIR = "/data/scratch/des/dr2_detection_output/run_1"
+
+
+def convert_to_global_idx(tile_idx, gpu_idx):
+    num_gpus = 2
+    tiles_per_img = 64
+    batch_size = 2
+    dir_idx = int(num_gpus * (tile_idx // (tiles_per_img / batch_size)) + gpu_idx)
+    subimage_idx = [(batch_size * tile_idx + i) % tiles_per_img for i in range(batch_size)]
+    return dir_idx, subimage_idx
+
+
+def convert_to_tile_idx(dir_idx):
+    num_gpus = 2
+    tiles_per_img = 64
+    batch_size = 2
+    gpu_idx = dir_idx % num_gpus
+    tile_starting_idx = (tiles_per_img / batch_size) * (dir_idx // num_gpus)
+    return int(tile_starting_idx), int(gpu_idx)
+
+
+def count_num_clusters(dir_idx):
+    memberships = torch.empty((0, 10, 10))
+    tile_starting_idx, gpu_idx = convert_to_tile_idx(dir_idx)
+    for tile in range(tile_starting_idx, tile_starting_idx + 32):
+        file = torch.load(
+            f"{OUTPUT_DIR}/rank_{gpu_idx}_batchIdx_{tile}_dataloaderIdx_0.pt",
+            map_location=torch.device("cpu"),
+        )
+        memberships = torch.cat((memberships, file["mode_cat"]["membership"].squeeze()), dim=0)
+    memberships = torch.repeat_interleave(memberships, repeats=128, dim=1)
+    memberships = torch.repeat_interleave(memberships, repeats=128, dim=2)
+    return torch.any(memberships.view(memberships.shape[0], -1), dim=1).sum()
+
+
+def main():
+    num_clusters = {}
+    output_filename = "/data/scratch/des/num_clusters.pickle"
+    for dir_idx, des_dir in enumerate(DES_SUBDIRS):
+        if dir_idx > 10167:
+            break
+        num_clusters[des_dir] = count_num_clusters(dir_idx)
+
+    with open(output_filename, "wb") as handle:
+        pickle.dump(num_clusters, handle, protocol=pickle.HIGHEST_PROTOCOL)
+
+
+if __name__ == "__main__":
+    main()