Tract-Patch Search

[1]:

from dask.distributed import Client
import lsdb
import skymap_convert
from lsdb_rubin import tract_patch_search

[2]:

client = Client(n_workers=4, memory_limit="auto")
client

[2]:

Client

Client-a4981894-a468-11f0-bb78-e4c767ffbe5f

Connection method: Cluster object	Cluster type: distributed.LocalCluster
Dashboard: http://127.0.0.1:8787/status

Cluster Info

LocalCluster

c1a11fd9

Dashboard: http://127.0.0.1:8787/status	Workers: 4
Total threads: 20	Total memory: 31.00 GiB
Status: running	Using processes: True

Scheduler Info

Scheduler

Scheduler-217aec54-4638-4aab-a748-1f0c3885601b

Comm: tcp://127.0.0.1:45321	Workers: 4
Dashboard: http://127.0.0.1:8787/status	Total threads: 20
Started: Just now	Total memory: 31.00 GiB

Workers

Worker: 0

Comm: tcp://127.0.0.1:41351	Total threads: 5
Dashboard: http://127.0.0.1:46781/status	Memory: 7.75 GiB
Nanny: tcp://127.0.0.1:43399
Local directory: /tmp/dask-scratch-space/worker-f00_hol8

Worker: 1

Comm: tcp://127.0.0.1:39261	Total threads: 5
Dashboard: http://127.0.0.1:41001/status	Memory: 7.75 GiB
Nanny: tcp://127.0.0.1:33149
Local directory: /tmp/dask-scratch-space/worker-6wmggf9i

Worker: 2

Comm: tcp://127.0.0.1:33105	Total threads: 5
Dashboard: http://127.0.0.1:45179/status	Memory: 7.75 GiB
Nanny: tcp://127.0.0.1:35163
Local directory: /tmp/dask-scratch-space/worker-12reep0v

Worker: 3

Comm: tcp://127.0.0.1:36093	Total threads: 5
Dashboard: http://127.0.0.1:46837/status	Memory: 7.75 GiB
Nanny: tcp://127.0.0.1:37231
Local directory: /tmp/dask-scratch-space/worker-6uf5h9sz

Load GAIA and the LSST skymap

The catalog starts with 2,000+ partitions, and we can see the distribution of those partitions on the sky with the below plot. The yellow areas are the galactic bulge, where there is higher angular density of objects, and so we use smaller data partitions in the HATS catalog.

[3]:

# Load GAIA DR3 data.

gaia = lsdb.open_catalog("https://data.lsdb.io/hats/gaia_dr3", columns=["ra", "dec", "source_id"])
gaia

[3]:

lsdb Catalog gaia:

	ra	dec	source_id
npartitions=2016
Order: 2, Pixel: 0	double[pyarrow]	double[pyarrow]	int64[pyarrow]
Order: 2, Pixel: 1	...	...	...
...	...	...	...
Order: 3, Pixel: 766	...	...	...
Order: 3, Pixel: 767	...	...	...

3 out of 152 available columns in the catalog have been loaded lazily, meaning no data has been read, only the catalog schema

[4]:

gaia.plot_pixels()

[4]:

(<Figure size 1000x500 with 2 Axes>,
 <WCSAxes: title={'center': 'Catalog pixel map - gaia'}>)

../_images/pre_executed_tract_patch_search_5_1.png

[5]:

# Load the LSST skymap.

import skymap_convert

lsst_skymap = skymap_convert.ConvertedSkymapReader(preset="lsst_skymap")

Search by tract ID

We search by a tract in the galactic bulge, and see that it overlaps with a few of the partitions in that region.

We can attach a new method called tract_patch_search, which is implemented to supplement other types of region search on the lsdb Catalog object. You can find more general information on region search in the LSDB documentation.

[6]:

from lsdb.core.plotting import plot_points
import astropy.units as u
from astropy.coordinates import SkyCoord

fov = (5 * u.deg, 5 * u.deg)
center = SkyCoord(265 * u.deg, -30 * u.deg)

tract_index = 4975
patch_index = 67

[7]:

# Only specify the tract (and not patch) to search by tract.

lsdb.catalog.Catalog.tract_patch_search = tract_patch_search

gaia_tract = gaia.tract_patch_search(skymap_reader=lsst_skymap, tract=tract_index)
gaia_tract

[7]:

lsdb Catalog gaia:

	ra	dec	source_id
npartitions=5
Order: 5, Pixel: 7206	double[pyarrow]	double[pyarrow]	int64[pyarrow]
Order: 5, Pixel: 7207	...	...	...
Order: 6, Pixel: 28848	...	...	...
Order: 6, Pixel: 28849	...	...	...
Order: 6, Pixel: 28852	...	...	...

3 out of 152 available columns in the catalog have been loaded lazily, meaning no data has been read, only the catalog schema

[8]:

gaia_tract.plot_pixels(
    plot_title="Gaia DR3 Tract Search",
    fc="#00000000",
    ec="red",
)

[8]:

(<Figure size 1000x500 with 2 Axes>,
 <WCSAxes: title={'center': 'Gaia DR3 Tract Search'}>)

../_images/pre_executed_tract_patch_search_10_1.png

Plot the points within our tract search

[9]:

df_tract = gaia_tract.compute()
df_tract

/home/delucchi/.virtualenvs/sep/lib/python3.12/site-packages/distributed/client.py:3383: UserWarning: Sending large graph of size 116.79 MiB.
This may cause some slowdown.
Consider loading the data with Dask directly
 or using futures or delayed objects to embed the data into the graph without repetition.
See also https://docs.dask.org/en/stable/best-practices.html#load-data-with-dask for more information.
  warnings.warn(

[9]:

	ra	dec	source_id
_healpix_29
2028429617617798730	266.228612	-29.751009	4056859222565665024
2028429623682899072	266.214002	-29.750143	4056859222553709184
...	...	...	...
2030314331564457470	265.403714	-28.265501	4060628657676158080
2030314331687825753	265.40231	-28.265104	4060628661956658816

395098 rows × 3 columns

Zooming in on the tract, we can see the ~400,000 points from the catalog, in a semi-rectangular “tract-shaped” region

[10]:

fig, ax = gaia.plot_pixels(projection="AIT", fov=fov, center=center, ec="000000")
_ = plot_points.plot_points(
    df_tract,
    ra_column="ra",
    dec_column="dec",
    title="Gaia DR3 Tract Search",
    fov=fov,
    center=center,
)

../_images/pre_executed_tract_patch_search_14_0.png

Search by patch ID

[11]:

# Specify both tract and patch to search by tract and patch.

gaia_tract_patch = gaia.tract_patch_search(
    skymap_reader=lsst_skymap,
    tract=tract_index,
    patch=patch_index,  # include patch here!
)
gaia_tract_patch

[11]:

lsdb Catalog gaia:

	ra	dec	source_id
npartitions=1
Order: 6, Pixel: 28849	double[pyarrow]	double[pyarrow]	int64[pyarrow]

3 out of 152 available columns in the catalog have been loaded lazily, meaning no data has been read, only the catalog schema

[12]:

gaia_tract_patch.plot_pixels(
    plot_title="Gaia DR3 Tract and Patch Search",
    fc="#00000000",
    ec="red",
)

[12]:

(<Figure size 1000x500 with 2 Axes>,
 <WCSAxes: title={'center': 'Gaia DR3 Tract and Patch Search'}>)

../_images/pre_executed_tract_patch_search_17_1.png

Plot the points within our patch search

[13]:

df_tract_patch = gaia_tract_patch.compute()
df_tract_patch

/home/delucchi/.virtualenvs/sep/lib/python3.12/site-packages/distributed/client.py:3383: UserWarning: Sending large graph of size 116.78 MiB.
This may cause some slowdown.
Consider loading the data with Dask directly
 or using futures or delayed objects to embed the data into the graph without repetition.
See also https://docs.dask.org/en/stable/best-practices.html#load-data-with-dask for more information.
  warnings.warn(

[13]:

	ra	dec	source_id
_healpix_29
2030076515776642061	265.294515	-28.840948	4060153015792358784
2030076516438355910	265.292557	-28.839182	4060153015802849152
...	...	...	...
2030095571649978731	265.299339	-28.674579	4060191125115382656
2030095571700128422	265.299429	-28.673891	4060191125095778944

10543 rows × 3 columns

Zooming in to the same sky area as above, we see only the ~10,000 points that are in the semi-rectangular patch region.

[14]:

fig, ax = gaia.plot_pixels(projection="AIT", fov=fov, center=center, ec="000000")
plot_points.plot_points(
    df_tract_patch,
    ra_column="ra",
    dec_column="dec",
    title="Gaia DR3 Tract and Patch Search",
    fov=fov,
    center=center,
)

[14]:

(<Figure size 1000x500 with 2 Axes>,
 <WCSAxes: title={'center': 'Gaia DR3 Tract and Patch Search'}>)

../_images/pre_executed_tract_patch_search_21_1.png

About

Authors: Olivia Lynn

Last updated on: October 8, 2025

If you use lsdb for published research, please cite following instructions.