def test_position_velocity(keep_figures=False):
with TESS_SPICE() as knl:
# We should be able to load and close without affecting the results of the following:
with TESS_SPICE():
pass
time_nocorr = np.array(
[1325.32351727, 1325.34435059, 1325.36518392, 1325.38601724])
# Get the location of TESS as a function of time relative to Earth in kilometers:
pos = knl.position(time_nocorr + 2457000)
assert pos.shape == (len(time_nocorr), 3)
# Get the location of TESS as a function of time relative to Earth in kilometers:
vel = knl.velocity(time_nocorr + 2457000)
assert vel.shape == (len(time_nocorr), 3)
# Plot TESS orbits in 3D:
time_inter = np.linspace(time_nocorr[0] - 3, time_nocorr[-1] + 3, 2000)
pos_inter = knl.position(time_inter + 2457000)
vel_inter = knl.velocity(time_inter + 2457000)
fig1 = plt.figure()
ax = fig1.add_subplot(111, projection='3d')
ax.plot(pos_inter[:, 0], pos_inter[:, 1], pos_inter[:, 2], 'r-')
ax.scatter(pos[:, 0], pos[:, 1], pos[:, 2], alpha=0.5)
ax.scatter(0, 0, 0, alpha=0.5, c='b')
fig2 = plt.figure()
ax1 = fig2.add_subplot(211)
ax1.plot(time_inter, np.linalg.norm(pos_inter, axis=1), 'r-')
ax1.scatter(time_nocorr, np.linalg.norm(pos, axis=1), alpha=0.5)
ax1.set_ylabel('Distance (km)')
ax1.set_xticks([])
ax2 = fig2.add_subplot(212)
ax2.plot(time_inter, np.linalg.norm(vel_inter, axis=1), 'r-')
ax2.scatter(time_nocorr, np.linalg.norm(vel, axis=1), alpha=0.5)
ax2.set_ylabel('Velocity (km/s)')
ax2.set_xlabel('Time')
plt.tight_layout()
if not keep_figures:
plt.close(fig1)
plt.close(fig2)
def test_sclk2jd():
print("=" * 72)
star_coord = coord.SkyCoord(ra='04:52:6.92',
dec='-70:43:52.4',
unit=(u.hourangle, u.deg),
frame='icrs',
pm_ra_cosdec=1.13036 * u.mas / u.yr,
pm_dec=-11.1042 * u.mas / u.yr,
obstime=Time("J2000"))
with TESS_SPICE() as knl:
desired = 1468.416666534158
jdtdb = knl.sclk2jd('1228946341.75')
time = jdtdb - 2457000
diff = (time - desired) * 86400
print("Converted time: %.16f" % time)
print("Desired: %.16f" % desired)
print("Difference: %.6f s" % diff)
#np.testing.assert_allclose(diff, 0, atol=0.1)
time, timecorr = knl.barycorr(jdtdb, star_coord)
time -= 2457000
diff = (time - desired) * 86400
print("Barycorr: %.6f" % (timecorr * 86400))
print("Converted time: %.16f" % time)
print("Difference: %.6f s" % diff)
print("=" * 72)
def test_spice(SHARED_INPUT_DIR, starid):
# Initialize our home-made TESS Kernel object:
with TESS_SPICE() as knl:
print("="*72)
print("TIC %d" % starid)
tpf_file = find_tpf_files(SHARED_INPUT_DIR, starid=starid)[0]
with fits.open(tpf_file, mode='readonly', memmap=True) as hdu:
time_tpf = hdu[1].data['TIME']
timecorr_tpf = hdu[1].data['TIMECORR']
camera = hdu[0].header['CAMERA']
ccd = hdu[0].header['CCD']
# Coordinates of the target as astropy SkyCoord object:
star_coord = coord.SkyCoord(
ra=hdu[0].header['RA_OBJ'],
dec=hdu[0].header['DEC_OBJ'],
unit=u.deg,
frame='icrs',
obstime=Time('J2000'),
pm_ra_cosdec=hdu[0].header['PMRA']*u.mas/u.yr,
pm_dec=hdu[0].header['PMDEC']*u.mas/u.yr,
radial_velocity=0*u.km/u.s
)
# Load the original timestamps from FFIs:
hdf_file = find_hdf5_files(SHARED_INPUT_DIR, camera=camera, ccd=ccd)[0]
with h5py.File(hdf_file, 'r') as hdf:
ffi_time = np.asarray(hdf['time'])
ffi_timecorr = np.asarray(hdf['timecorr'])
f = interp1d(time_tpf-timecorr_tpf, timecorr_tpf, kind='linear')
# Change the timestamps bach to JD:
time_nocorr = ffi_time - ffi_timecorr
times = Time(time_nocorr, 2457000, format='jd', scale='tdb')
ras, decs = add_proper_motion(
star_coord.ra.value,
star_coord.dec.value,
star_coord.pm_ra_cosdec.value,
star_coord.pm_dec.value,
times.jd
)
star_coord = coord.SkyCoord(
ra=ras[0],
dec=decs[0],
unit=u.deg,
frame='icrs'
)
print(star_coord)
# Use Greenwich as location instead of TESS (similar to what is done in Elenor):
greenwich = coord.EarthLocation.of_site('greenwich')
times_greenwich = Time(time_nocorr+2457000, format='jd', scale='utc', location=greenwich)
timecorr_greenwich = times_greenwich.light_travel_time(star_coord, kind='barycentric', ephemeris='builtin').value
#time_greenwich = time_nocorr + timecorr_greenwich
# Calculate barycentric correction using our method:
time_astropy, timecorr_astropy = knl.barycorr(time_nocorr + 2457000, star_coord)
# Calculate barycentric correction using second method:
timecorr_knl = knl.barycorr2(time_nocorr + 2457000, star_coord)
print(timecorr_knl)
# Plot the new barycentric time correction and the old one:
fig1 = plt.figure()
ax = fig1.add_subplot(111)
ax.plot(time_tpf - timecorr_tpf, timecorr_tpf*86400, 'g.-', label='TPF timecorr')
ax.plot(time_nocorr, ffi_timecorr*86400, 'b.-', label='FFI centre')
ax.scatter(time_nocorr, timecorr_greenwich*86400, marker='x', label='Elenor timecorr')
ax.scatter(time_nocorr, timecorr_astropy*86400, marker='x', c='r', label='Our timecorr')
ax.set_xlabel('Uncorrected Time [TJD]')
ax.set_ylabel('Barycentric Time Correction [s]')
ax.set_title('TIC %d' % starid)
ax.set_yscale('log')
plt.legend()
# Plot the new barycentric time correction and the old one:
fig2 = plt.figure(figsize=(8,10))
ax3 = fig2.add_subplot(111)
ax3.axhline(1.0, color='k', ls='--', lw=1)
ax3.plot(time_nocorr, np.abs(ffi_timecorr - f(time_nocorr))*86400, '.-', label='FFI centre')
ax3.plot(time_nocorr, np.abs(timecorr_greenwich - f(time_nocorr))*86400*1000, '.-', label='Elenor')
ax3.plot(time_nocorr, np.abs(timecorr_astropy - f(time_nocorr))*86400*1000, '.-', label='Astropy')
ax3.plot(time_nocorr, np.abs(timecorr_knl - f(time_nocorr))*86400*1000, '.-', label='Kernel')
ax3.set_ylabel('abs(TimeCorr - TPF TimeCorr) [ms]')
ax3.set_xlabel('Uncorrected Time [TJD]')
ax3.set_yscale('log')
ax3.set_ylim(bottom=0.1)
ax3.legend()
ax3.grid(axis='y', which='both', color='0.7', ls=':', lw=0.5)
ax3.set_title('TIC %d' % starid)
plt.tight_layout()
assert np.all(86400000*np.abs(timecorr_astropy - f(time_nocorr)) < 1), "AstroPy method: More than 1 ms away"
assert np.all(86400000*np.abs(timecorr_knl - f(time_nocorr)) < 1), "HomeMade method: More than 1 ms away"
print("="*72)
def test_spice(keep_figures=False):
# Initialize our home-made TESS Kernel object:
with TESS_SPICE() as knl:
for starid in (260795451, 267211065):
print("=" * 72)
print("TIC %d" % starid)
tpf_file = find_tpf_files(INPUT_DIR, starid=starid)[0]
with fits.open(tpf_file, mode='readonly', memmap=True) as hdu:
time_tpf = hdu[1].data['TIME']
timecorr_tpf = hdu[1].data['TIMECORR']
camera = hdu[0].header['CAMERA']
ccd = hdu[0].header['CCD']
# Coordinates of the target as astropy SkyCoord object:
star_coord = coord.SkyCoord(
ra=hdu[0].header['RA_OBJ'],
dec=hdu[0].header['DEC_OBJ'],
unit=u.deg,
frame='icrs',
obstime=Time('J2000'),
pm_ra_cosdec=hdu[0].header['PMRA'] * u.mas / u.yr,
pm_dec=hdu[0].header['PMDEC'] * u.mas / u.yr,
radial_velocity=0 * u.km / u.s)
# Load the original timestamps from FFIs:
hdf_file = find_hdf5_files(INPUT_DIR, camera=camera, ccd=ccd)[0]
with h5py.File(hdf_file, 'r') as hdf:
ffi_time = np.asarray(hdf['time'])
ffi_timecorr = np.asarray(hdf['timecorr'])
f = interp1d(time_tpf - timecorr_tpf, timecorr_tpf, kind='linear')
# Change the timestamps bach to JD:
time_nocorr = ffi_time - ffi_timecorr
times = Time(time_nocorr, 2457000, format='jd', scale='tdb')
ras, decs = add_proper_motion(star_coord.ra.value,
star_coord.dec.value,
star_coord.pm_ra_cosdec.value,
star_coord.pm_dec.value, times.jd)
star_coord = coord.SkyCoord(ra=ras[0],
dec=decs[0],
unit=u.deg,
frame='icrs')
print(star_coord)
# Use Greenwich as location instead of TESS (similar to what is done in Elenor):
greenwich = coord.EarthLocation.of_site('greenwich')
times_greenwich = Time(time_nocorr + 2457000,
format='jd',
scale='utc',
location=greenwich)
timecorr_greenwich = times_greenwich.light_travel_time(
star_coord, kind='barycentric', ephemeris='builtin').value
#time_greenwich = time_nocorr + timecorr_greenwich
# Calculate barycentric correction using our method:
time_astropy, timecorr_astropy = knl.barycorr(
time_nocorr + 2457000, star_coord)
# Caluclate barycentric correction uning second method:
timecorr_knl = knl.barycorr2(time_nocorr + 2457000, star_coord)
print(timecorr_knl)
# Plot the new barycentric time correction and the old one:
fig1 = plt.figure()
ax = fig1.add_subplot(111)
ax.scatter(time_nocorr,
ffi_timecorr * 86400,
alpha=0.3,
s=4,
label='FFI timecorr')
ax.scatter(time_tpf - timecorr_tpf,
timecorr_tpf * 86400,
alpha=0.3,
s=4,
label='TPF timecorr')
ax.scatter(time_nocorr,
timecorr_greenwich * 86400,
alpha=0.3,
s=4,
label='Elenor timecorr')
ax.scatter(time_nocorr,
timecorr_astropy * 86400,
alpha=0.3,
s=4,
label='Our timecorr')
ax.set_xlabel('Uncorrected Time (JD - 2457000)')
ax.set_ylabel('Barycentric Time Correction (s)')
ax.set_title('TIC %d' % starid)
plt.legend()
# Plot the new barycentric time correction and the old one:
fig2 = plt.figure(figsize=(8, 10))
ax1 = fig2.add_subplot(311)
ax1.axhline(0, color='k', ls=':', lw=0.5)
ax1.plot(time_nocorr, (ffi_timecorr - f(time_nocorr)) * 86400, '.')
ax1.set_ylabel('FFI - TPF (seconds)')
ax1.set_title('TIC %d' % starid)
ax2 = fig2.add_subplot(312)
ax2.axhline(0, color='k', ls=':', lw=0.5)
ax2.plot(time_nocorr,
(timecorr_greenwich - f(time_nocorr)) * 86400 * 1000, '.')
ax2.set_ylabel('Elenor - TPF (ms)')
ax3 = plt.subplot(313)
ax3.axhline(0, color='k', ls=':', lw=0.5)
ax3.plot(time_nocorr,
(timecorr_astropy - f(time_nocorr)) * 86400 * 1000,
'.',
label='Astropy')
ax3.plot(time_nocorr,
(timecorr_knl - f(time_nocorr)) * 86400 * 1000,
'.',
label='Kernel')
ax3.set_ylabel('Our - TPF (ms)')
ax3.set_xlabel('Uncorrected Time (TJD)')
ax3.legend()
ax1.set_xticks([])
ax2.set_xticks([])
plt.tight_layout()
if not keep_figures:
plt.close(fig1)
plt.close(fig2)
print("=" * 72)
[
38,
datetime.strptime('04/28/21', '%m/%d/%y'),
datetime.strptime('05/26/21', '%m/%d/%y')
],
[
39,
datetime.strptime('05/26/21', '%m/%d/%y'),
datetime.strptime('06/24/21', '%m/%d/%y')
],
]
fig = plt.figure(figsize=(15, 6), dpi=100)
ax = fig.add_subplot(111)
with TESS_SPICE() as ts:
# This requires the "brief" utility tool
# https://naif.jpl.nasa.gov/naif/utilities.html
# TODO: There is proberly a way to do this with SpiceyPy
proc = subprocess.Popen('brief --95 -utc "' + ts.METAKERNEL + '"',
shell=True,
stdout=subprocess.PIPE,
universal_newlines=True)
stdout, stderr = proc.communicate()
lines = stdout.split("\n")
k = -1
for line in lines:
line = line.strip()
if not line: continue
if line.startswith('Summary for: '):
plt.rc('text', usetex=True)
if __name__ == '__main__':
plt.switch_backend('Qt5Agg')
plt.close('all')
INPUT_DIR = r'F:\tess_data\S06_DR08'
fig4 = plt.figure()
ax = fig4.add_subplot(111)
ax.axhline(0, color='k', ls='--')
for starid in (59545062, 25155664, 14003429):
# Initialize our home-made TESS Kernel object:
with TESS_SPICE() as knl:
print("=" * 72)
print("TIC %d" % starid)
tpf_file = find_tpf_files(INPUT_DIR, starid=starid)[0]
with fits.open(tpf_file, mode='readonly', memmap=True) as hdu:
time_tpf = hdu[1].data['TIME']
timecorr_tpf = hdu[1].data['TIMECORR']
camera = hdu[0].header['CAMERA']
ccd = hdu[0].header['CCD']
print(camera, ccd)
# Coordinates of the target as astropy SkyCoord object:
star_coord = coord.SkyCoord(
ra=hdu[0].header['RA_OBJ'],
dec=hdu[0].header['DEC_OBJ'],