Use default_columns property from hats catalog if available in read_hats

src/lsdb/loaders/hats/read_hats.py

@@ -61,6 +61,12 @@ def read_hats(
         Catalog object loaded from the given parameters
     """
     # Creates a config object to store loading parameters from all keyword arguments.
+
+    hc_catalog = hc.read_hats(path)
+
+    if columns is None and hc_catalog.catalog_info.default_columns is not None:
+        columns = hc_catalog.catalog_info.default_columns
+
     config = HatsLoadingConfig(
         search_filter=search_filter,
         columns=columns,
@@ -69,7 +75,6 @@ def read_hats(
         kwargs=kwargs,
     )
 
-    hc_catalog = hc.read_hats(path)
     if hc_catalog.schema is None:
         raise ValueError(
             "The catalog schema could not be loaded from metadata."

tests/conftest.py

@@ -20,6 +20,7 @@
 SMALL_SKY_TO_XMATCH_SOFT_NAME = "small_sky_to_xmatch_soft"
 SMALL_SKY_ORDER1_DIR_NAME = "small_sky_order1"
 SMALL_SKY_ORDER1_NO_PANDAS_DIR_NAME = "small_sky_order1_no_pandas_meta"
+SMALL_SKY_ORDER1_DEFAULT_COLS_DIR_NAME = "small_sky_order1_default_columns"
 SMALL_SKY_ORDER1_SOURCE_NAME = "small_sky_order1_source"
 SMALL_SKY_ORDER1_SOURCE_MARGIN_NAME = "small_sky_order1_source_margin"
 SMALL_SKY_TO_ORDER1_SOURCE_NAME = "small_sky_to_o1source"
@@ -76,6 +77,11 @@ def small_sky_order1_dir(test_data_dir):
     return test_data_dir / SMALL_SKY_ORDER1_DIR_NAME
 
 
+@pytest.fixture
+def small_sky_order1_default_cols_dir(test_data_dir):
+    return test_data_dir / SMALL_SKY_ORDER1_DEFAULT_COLS_DIR_NAME
+
+
 @pytest.fixture
 def small_sky_order1_no_pandas_dir(test_data_dir):
     return test_data_dir / SMALL_SKY_ORDER1_NO_PANDAS_DIR_NAME

tests/data/generate_data.ipynb

@@ -123,6 +123,45 @@
     "sso1.partition_info.write_to_file(hats.io.paths.get_partition_info_pointer(out_catalog_name))"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### small_sky_order1_default_columns\n",
+    "\n",
+    "Copies small_sky_order1 but adds a list of default columns to the properties file"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2025-01-23T18:41:22.261748Z",
+     "start_time": "2025-01-23T18:41:22.230037Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "import os\n",
+    "import hats\n",
+    "from hats.io.paths import DATASET_DIR\n",
+    "from hats.io.file_io import get_upath\n",
+    "\n",
+    "out_catalog_name = \"small_sky_order1_default_columns\"\n",
+    "\n",
+    "sso1 = hats.read_hats(\"small_sky_order1\")\n",
+    "sso1_dataset_path = get_upath(\"small_sky_order1\") / DATASET_DIR\n",
+    "out_dataset_path = get_upath(out_catalog_name) / DATASET_DIR\n",
+    "\n",
+    "os.mkdir(out_catalog_name)\n",
+    "os.symlink(f\"../{sso1_dataset_path}\", out_dataset_path)\n",
+    "sso1.catalog_info.copy_and_update(\n",
+    "    catalog_name=out_catalog_name, default_columns=[\"ra\", \"dec\", \"id\"]\n",
+    ").to_properties_file(out_catalog_name)\n",
+    "sso1.partition_info.write_to_file(hats.io.paths.get_partition_info_pointer(out_catalog_name))"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},

tests/data/small_sky_order1_default_columns/dataset

@@ -0,0 +1 @@
+../small_sky_order1/dataset

tests/data/small_sky_order1_default_columns/partition_info.csv

@@ -0,0 +1,5 @@
+Norder,Npix
+1,44
+1,45
+1,46
+1,47

tests/data/small_sky_order1_default_columns/properties

@@ -0,0 +1,15 @@
+#HATS catalog
+obs_collection=small_sky_order1_default_columns
+dataproduct_type=object
+hats_nrows=131
+hats_col_ra=ra
+hats_col_dec=dec
+hats_cols_default=ra dec id
+hats_max_rows=1000000
+hats_order=1
+moc_sky_fraction=0.08333
+hats_builder=hats-import v0.3.6.dev26+g40366b4
+hats_creation_date=2024-10-15T14\:47UTC
+hats_estsize=39
+hats_release_date=2024-09-18
+hats_version=v0.1

tests/lsdb/loaders/hats/test_read_hats.py

@@ -38,6 +38,31 @@ def test_read_hats(small_sky_order1_dir, small_sky_order1_hats_catalog, assert_d
     assert_index_correct(catalog)
 
 
+def test_read_hats_default_cols(small_sky_order1_default_cols_dir, assert_divisions_are_correct):
+    catalog = lsdb.read_hats(small_sky_order1_default_cols_dir)
+    assert isinstance(catalog, lsdb.Catalog)
+    assert isinstance(catalog._ddf, nd.NestedFrame)
+    assert catalog.hc_structure.catalog_info.default_columns is not None
+    assert np.all(catalog.columns == catalog.hc_structure.catalog_info.default_columns)
+    assert np.all(catalog.compute().columns == catalog.hc_structure.catalog_info.default_columns)
+    assert isinstance(catalog.compute(), npd.NestedFrame)
+    assert_divisions_are_correct(catalog)
+    assert_index_correct(catalog)
+
+
+def test_read_hats_default_cols_specify_cols(small_sky_order1_default_cols_dir, assert_divisions_are_correct):
+    filter_columns = ["ra", "dec"]
+    catalog = lsdb.read_hats(small_sky_order1_default_cols_dir, columns=filter_columns)
+    assert isinstance(catalog, lsdb.Catalog)
+    assert isinstance(catalog._ddf, nd.NestedFrame)
+    assert catalog.hc_structure.catalog_info.default_columns is not None
+    assert np.all(catalog.columns == filter_columns)
+    assert np.all(catalog.compute().columns == filter_columns)
+    assert isinstance(catalog.compute(), npd.NestedFrame)
+    assert_divisions_are_correct(catalog)
+    assert_index_correct(catalog)
+
+
 def test_read_hats_no_pandas(small_sky_order1_no_pandas_dir, assert_divisions_are_correct):
     catalog = lsdb.read_hats(small_sky_order1_no_pandas_dir)
     assert isinstance(catalog, lsdb.Catalog)