Structured Search

A structured search on an LSDB catalog has two components:

• a coarse filter to limit which data partitions will be loaded

• a fine filter to limit the rows inside the data partitions

Most commonly, these are region-based filters, since the partitions can be easily determined for specific regions on the sphere. Culling data according to some shared region is a good way to limit the data you’re loading, while prototyping some full-sky pipeline.

There are a few ways to execute a structured search on LSDB catalogs.

The preferred approach is at the time of loading a catalog. You can instantiate some search_object and pass to the open call, with something like: lsdb.open_catalog(path, search_filter=search_object)

We provide several built-in search types:

ConeSearch(ra, dec, radius_arcsec[, fine])	Perform a cone search to filter the catalog
BoxSearch(ra, dec[, fine])	Perform a box search to filter the catalog.
PolygonSearch(vertices[, fine])	Perform a polygonal search to filter the catalog.
OrderSearch([min_order, max_order])	Filter the catalog by HEALPix order.
PixelSearch(pixels)	Filter the catalog by HEALPix pixels.
MOCSearch(moc[, fine])	Filter the catalog by a MOC.

IndexSearch(values, index_catalogs[, fine])

Find rows by column values using HATS index catalogs.

Alternatively, you can call the search method after a catalog has been opened and other operations have been performed with catalog.search(search_object). You can also call region search methods on the catalog object.

To see region searches in action, check out Region Selection.

To see an example of defining your own type of structured search, see Creating a custom search with global statistics.

Catalog.cone_search(ra, dec, radius_arcsec)	Perform a cone search to filter the catalog.
Catalog.box_search(ra, dec[, fine])	Performs filtering according to right ascension and declination ranges.
Catalog.polygon_search(vertices[, fine])	Perform a polygonal search to filter the catalog.
Catalog.id_search(values[, index_catalogs, fine])	Query rows by column values.
Catalog.order_search([min_order, max_order])	Filter catalog by order of HEALPix.
Catalog.pixel_search(pixels)	Finds all catalog pixels that overlap with the requested pixel set.
Catalog.moc_search(moc[, fine])	Finds all catalog points that are contained within a moc.
Catalog.search(search)	Find rows by reusable search algorithm.