Merge pull request #118 from astronomy-commons/sandro/update-crossmat…

…ch-docs Update custom crossmatch function docs
astronomy-commons · Jan 11, 2024 · e265018 · e265018
2 parents 20c2d67 + cb7fa01
commit e265018
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 8 deletions.
diff --git a/src/lsdb/catalog/catalog.py b/src/lsdb/catalog/catalog.py
@@ -132,9 +132,13 @@ def crossmatch(
 
                     The function should be able to perform a crossmatch on two pandas DataFrames
                     from a HEALPix pixel from each catalog. It should return a dataframe with the
-                    combined set of columns from the input dataframes with the appropriate suffixes,
-                    and a column with the name {AbstractCrossmatchAlgorithm.DISTANCE_COLUMN_NAME}
-                    with the distance between the points.
+                    combined set of columns from the input dataframes with the appropriate suffixes
+                    and, eventually, a set of extra columns generated by the crossmatch algorithm.
+                    These columns are specified in {AbstractCrossmatchAlgorithm.extra_columns}, with
+                    their respective data types, by means of an empty pandas dataframe. As an example,
+                    the KdTreeCrossmatch algorithm outputs a "_DIST" column with the distance between
+                    data points. Its extra_columns attribute is specified as follows:
+                    `pd.DataFrame({"_DIST": pd.Series(dtype=np.dtype("float64"))})`.
 
                     The class will have been initialized with the following parameters, which the
                     crossmatch function should use:
@@ -161,9 +165,8 @@ def crossmatch(
             pair of neighbors found from cross-matching.
 
             The resulting table contains all columns from the left and right catalogs with their
-            respective suffixes, and a column with the name
-            {AbstractCrossmatchAlgorithm.DISTANCE_COLUMN_NAME} with the great circle separation
-            between the points.
+            respective suffixes and, whenever specified, a set of extra columns generated by the
+            crossmatch algorithm.
         """
         if suffixes is None:
             suffixes = (f"_{self.name}", f"_{other.name}")

diff --git a/src/lsdb/core/crossmatch/kdtree_match.py b/src/lsdb/core/crossmatch/kdtree_match.py
@@ -33,8 +33,7 @@ def crossmatch(
             A DataFrame from the left and right tables merged with one row for each pair of
             neighbors found from cross-matching. The resulting table contains the columns from the
             left table with the first suffix appended, the right columns with the second suffix, and
-            a column with the name {AbstractCrossmatchAlgorithm.DISTANCE_COLUMN_NAME} with the great
-            circle separation between the points.
+            a "_DIST" column with the great circle separation between the points.
         """
         # Distance in 3-D space for unit sphere
         d_chord = 2.0 * math.sin(math.radians(0.5 * d_thresh))