ztffields documentation#
ztffields is a package made to handle Zwicky Transient Facility (ZTF) fields system.
ZTF observes using pre-define sets of field idenfitied with a unique fieldid ; it already exists more than 2000 fieldids.
The ZTF camera is made of 16 CCDs identified by a unique ccdid ranging from 1 to 16 (incl.) and each CCD has four amplifier, definining 4 quadrants defining, in total 64 rcid (16x4) ranging from 0 to 63 (incl.).
ztffields enables to interact with the ztf field system at the focalplane (whole camera footprint ; 1 polygon), at the ccd (16 polygons) or are the quadrant (64 polygons) levels.
Sharp start#
Get the fieldid that contains a given set of targets
import pandas
import numpy as np
import ztffields
np.random.seed(1234) # seed
# generate 40_000 RA, Dec coordinates with some bounds
size = 40_000
dec_range = [-45, 90]
ra_range=[0,360]
# -> random in the sky
dec_sin_range = np.sin(np.asarray(dec_range)*np.pi/180)
ra = np.random.uniform(*ra_range, size=size)
dec = np.arcsin( np.random.uniform(*dec_sin_range, size=size) ) / (np.pi/180)
# build a dataframe structure.
radecs = pandas.DataFrame({"ra":ra, "dec":dec})
# and project to a focalplane shape | try using ccd or quadrant
df = ztffields.radec_to_fieldid(radecs, level="focalplane")
df.loc[600] # see the fieldid containing the target 600
# result
index_radec
600 1729
600 688
Name: fieldid, dtype: object
- Performance wise, on a regular laptop, matching 10_000 targets takes:
300ms at the focalplane level)
1.7s at the ccd level
4.8s at the quadrant level
Now display the number of target per main grid fields
# groupby field id
fieldid_s = df.to_frame().groupby("fieldid").size()
# - and limit to the main grid
fieldid_s = fieldid_s[fieldid_s.index<1000]
# Display
fig = ztffields.skyplot_fields(fieldid_s)
Tweack how to display
# change the plotting system,
# add a label
# change the colormap
# using a cartopy projection
import cartopy.crs as ccrs
fig = ztffields.skyplot_fields(fieldid_s,
system="galactic",
label="number of observations",
cmap="inferno_r",
projection=ccrs.Robinson())
Documentation#
Code documentation