comments

utils/keck_filtering.ipynb

@@ -28,6 +28,12 @@
     "import numpy as np\n",
     "import shutil\n",
     "\n",
+    "\"\"\"\n",
+    "Use this code after downloading imagery using\n",
+    "keck_downloading file.\n",
+    "\n",
+    "\"\"\"\n",
+    "\n",
     "def get_all_fits_files(root_dir):\n",
     "    # Use glob to recursively find all .fits files\n",
     "    pattern = os.path.join(root_dir, '**', '*LR*.fits')\n",
@@ -103,6 +109,12 @@
     "latitudes = list(df['dec'])\n",
     "longitudes = list(df['ra'])\n",
     "\n",
+    "\"\"\"\n",
+    "Code to compute all angular separations between pairwise images from single RA DEC\n",
+    "values.\n",
+    "\n",
+    "\"\"\"\n",
+    "\n",
     "n_points = len(latitudes)\n",
     "\n",
     "# Repeat each point n_points times for lat1, lon1\n",
@@ -160,6 +172,11 @@
     "KECK_FOV = 3768 * KECK_DEG_PER_PIXEL\n",
     "THRESH = KECK_FOV * 2\n",
     "\n",
+    "'''\n",
+    "Initial agglomerative clustering.\n",
+    "Since we don't have WCS info, the above threshold is very conservative.\n",
+    "'''\n",
+    "\n",
     "clustering = AgglomerativeClustering(n_clusters=None, metric='precomputed', linkage='single', distance_threshold=THRESH)\n",
     "labels = clustering.fit_predict(angular_separations_matrix)"
    ]
@@ -196,6 +213,11 @@
     "RA_NAME = 'ra'\n",
     "DEC_NAME = 'dec'\n",
     "\n",
+    "\"\"\"\n",
+    "Only select images that are at least THRESH apart from each other.\n",
+    "\n",
+    "\"\"\"\n",
+    "\n",
     "def max_subset_with_min_distance(points, min_distance):\n",
     "    subset = []\n",
     "    for i, row in points.iterrows():\n",