Tools to organize and query astronomical catalogs
tools to organize and query astronomical catalogs
This modules provides classes to import astronomical catalogs into a mongodb database, and to efficiently query this database for positional matches.
The two main classes of this module are:
CatalogPusher: will process the raw files with the catalog sources and creates a database. See insert_example notebook for more details and usage instruction.
CatalogQuery: will perform queries on the catalogs. See query_example for examples and benchmarking.
Supported queries includes:
all the sources with a certain distance.
closest source at a given position.
binary search: return yes/no if anything is around the positon.
user defined queries.
The first item on the above list (cone search around target) provides the basic block for the other two types of positional-based queries. The code supports tree types of basic cone-search queries, depending on the indexing strategy of the database.
using HEALPix: if the catalog sources have been assigned an HEALPix index (using healpy).
using GeoJSON (or ‘legacy coordinates’): if the catalog documents have the position arranged in one of these two formats (example), the query is based on the $geoWithin and $centerSphere mongo operators.
All the core functions are defined in the catquery_utils module. In all cases the results of the queries will be return an astropy.table.Table objects.
Notes on indexing and query performances:
The recommended method to index and query catalogs is based on the GeoJSON coorinate type. See the example_insert notebook for how this can be implemented.
Performant queries requires the database indexes to reside in the RAM. The indexes are efficiently compressed by mongodb default engine (WiredTiger), however there is little redundant (and hence compressible) information in accurately measured coordinate pairs. As a consequence, GeoJSON type indexes tends to require fair amount of free memory (of the order 40 MB for 2M entries). For large catalogs (and / or small RAM) indexing on coordinates might not be feasible. In this case, the HEALPix based indexing should be used. As (possibly) many sources shares the same HEALPix index, compression is more efficient into moderating RAM usage.
The easiest way to install the Python library is with pip:
pip install extcats
If you want do modify extcats itself, you’ll need an editable installation. After cloning this Git repository:
Release history Release notifications | RSS feed
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.