zort : ZTF Object Reader Tool
The ZTF Object Reader Tool,
zort, is set of functions to organize and
access the ZTF Public Data Release lightcurves across multiple colors.
ZTF Public Data Release Lightcurves
Instructions for downloading and extracting ZTF Public Data Release Lightcurves can be found at: https://www.ztf.caltech.edu/page/dr2#12c
The ZTF Public Data Release Lightcurves are generated through spatially cross-matching individual epoch photometric catalogs. Catalogs are pre-filtered to be (1) the same ZTF observation field ID, (2) the same CCD readout channel, and (3) the same photometric color. Spatially coincidence observations in these catalogs are all labelled as objects and saved to a common ascii file along with the observation data for each epoch of the object. These files are consolidated such that all objects sharing a common ZTF observation field ID reside in the same file.
zort refers to these files with extension
zort provides facilitates the reading and inspection of lightcurves in
the ZTF Public Data Release. The features of
- Seamless looping through ZTF lightcurves for custom filtering, where interesting objects can be saved and recovered by only their file location
- Consolidating g-band and R-band lightcurves of a single source that are otherwise labelled as two separate objects by pairing objects as "siblings"
- Plotting lightcurves in multiple colors for visual inspection
Preferred method is through pip:
pip install zort
Latest version can also be installed from github:
git clone https://github.com/MichaelMedford/zort.git cd zort python setup.py install
- lightcurve file: Files included in the ZTF Public Data Release containing epoch photometry for spatially coincidence observations
- object: A collection of spatially coincident observations in a single color. Objects include IDs, sky locations (in right ascension and declination) and colors (g-band and R-band).
- lightcurve: Observation epochs of an object. Lightcurve observations include dates, magnitudes and magnitude errors.
- radec_map: Binary search trees for the objects in a lightcurve file. required for faster object access.
- siblings: A spatially coincident object in a different color originating from the same astrophysical source.
zort requires two additional data products per lightcurve file
*.txt) in order to make object discovery and multiple color
consolidation faster. Object files (
*.objects) contain all of the
metadata for each object in a lightcurve file. RCID map files
*.radec_map) contain binary search trees that facilitates faster
matching of multiple colors for individual objects.
zort requires that
each lightcurve file has a corresponding object file and RCID map file.
To generate object files and RCID map files for a directory of lightcurve files, run
zort-initialize -lightcurve-file-directory=LIGHTCURVE_FILE_DIRECTORY -single
or if mpi4py is installed then launch multiple instances of
zort-initialize -lightcurve-file-directory=LIGHTCURVE_FILE_DIRECTORY -parallel
If each lightcurve file does not have an object file and an RCID map then
zort will not be able to locate siblings
zort is designed to provide you with easy access to all of the
lightcurves in a lightcurve file for applying filters and saving interesting
objects. The preferred method for inspecting lightcurves is through a for-loop.
A filter is created that returns True for interesting objects. This filter can involve simply cuts on object properties or complicated model fitting to the full observation data in the object's lightcurve
def my_interesting_filter(obj): cond1 = obj.nepochs >= 20 cond2 = min(obj.lightcurve.mag) <= 17.0 if cond1 and cond2: return True else: return False
When a lightcurve file is looped over, it returns each object in the lightcurve
file. Interesting objects can be gathered into a list and saved to disk using the
filename = 'lightcurve_file_example.txt' interesting_objects =  from zort.lightcurveFile import LightcurveFile for obj in LightcurveFile(filename): if my_interesting_filter(obj): interesting_objects.append(obj) from zort.object import save_objects save_objects('objects.list', interesting_objects)
Objects and their lightcurves can be retrieved from a saved list by using the
load_objects function. Each object comes loaded with its metadata and
lightcurve, easily previewed by printing the object and lightcurve attribute.
from zort.object import load_objects interesting_objects = load_objects('objects.list') for obj in interesting_objects: print(obj) print(obj.lightcurve)
Objects can also be extracted in parallel by instantiating the LightcurveFile class with a rank and size. This could be done through mpi4py, or other parallelization packages. The LightcurveFile class simply needs to be told the rank of the parallel process and the total number, or size, of the parallel processes.
from mpi4py import MPI comm = MPI.COMM_WORLD rank = comm.rank size = comm.size filename = 'lightcurve_file_example.txt' interesting_objects =  from zort.lightcurveFile import LightcurveFile for obj in LightcurveFile(filename, proc_rank=rank, proc_size=size): if my_interesting_filter(obj): interesting_objects.append(obj) from zort.object import save_objects save_objects('objects.%i.list' % rank, interesting_objects)
proc_size parameters will cause the
iterator to uniquely send different objects to each parallel process without
loading all of the objects into memory for each process. This allows for
applying a filter to all of the objects in a lightcurve file without
Matching multiple colors for an object
Each object is defined as a spatially coincidence series of observations that share a (1) ZTF observation field ID, (2) CCD readout channel, and (3) photometric filter. This labels multiple colors of the same astrophysical source as separate ZTF objects with separate object IDs. The ZTF Public Data Release does not provide any native support for pairing these objects as multiple colors of the same source.
zort supports searching for and saving multiple colors for the same
source. The ZTF Public Data Release contains observations in g-band
(filterid=1) and R-band (filterid=2). Each object can therefore have one
additional object that comes from the same astrophysical source but is in a
different color. These matching objects are labelled as "siblings" and can
be both discovered and saved with
The siblings for each object can be located by simply running an object's
locate_siblings method. Running
filename = 'field000245_ra357.03053to5.26702_dec-27.96964to-20.4773.txt' buffer_position = 6852 obj = Object(filename, buffer_position) obj.locate_siblings(printFlag=True)
Locating siblings for ZTF Object 245101100000025 -- Object location: 4.74852, -26.23583 ... ** siblings file missing! ** -- Searching between buffers 17749819 and 18135260 ---- Sibling found at 4.74851, -26.23581 ! ---- Original Color: 1 | Sibling Color: 2 ---- Sibling saved
An object's siblings is itself another object and can be accessed through the siblings attribute.
print(obj) Filename: field000245_ra357.03053to5.26702_dec-27.96964to-20.4773.txt Buffer Position: 6852 Object ID: 245101100000025 Color: g Ra/Dec: (4.74852, -26.23583) 22 Epochs passing quality cuts print(obj.siblings) Filename: field000245_ra357.03053to5.26702_dec-27.96964to-20.4773.txt Buffer Position: 126136890 Object ID: 245201100000047 Color: r Ra/Dec: (4.74851, -26.23581) 22 Epochs passing quality cuts
The default tolerance for matching two objects as siblings is 2.0". However
this can be altered by setting the
- python 3.6
- Michael Medford MichaelMedford@berkeley.edu