def gmst_0h(gps):
"""
Truncate a LIGOTimeGPS to Greenwich mean sidereal 0 rad.
"""
gmst = lal.GreenwichMeanSiderealTime(gps)
residual = gmst % (2.0 * math.pi)
if residual:
gmst -= residual
return lal.GreenwichMeanSiderealTimeToGPS(gmst)
def pick_coinc():
"""Pick a coincidence from the "First Two Years" paper."""
filename = pkg_resources.resource_filename(
__name__, '../data/first2years/2016/gstlal.xml.gz')
xmldoc = utils.load_filename(filename, contenthandler=ContentHandler)
root, = xmldoc.childNodes
# Remove unneeded tables
for lsctable in (lsctables.FilterTable, lsctables.SegmentTable,
lsctables.SegmentDefTable, lsctables.SimInspiralTable,
lsctables.SummValueTable, lsctables.SearchSummVarsTable):
root.removeChild(lsctable.get_table(xmldoc))
coinc_inspiral_table = table = lsctables.CoincInspiralTable.get_table(
xmldoc)
# Determine event with most recent sideral time
gps_time_now = lal.GPSTimeNow()
gmsts = np.asarray([lal.GreenwichMeanSiderealTime(_.end) for _ in table])
gmst_now = lal.GreenwichMeanSiderealTime(gps_time_now)
div, rem = divmod(gmst_now - gmsts, 2 * np.pi)
i = np.argmin(rem)
new_gmst = div[i] * 2 * np.pi + gmsts[i]
old_time = table[i].end
new_time = lal.LIGOTimeGPS()
result = lal.GreenwichMeanSiderealTimeToGPS(new_gmst, new_time)
result.disown()
del result
delta_t = new_time - old_time
target_coinc_event_id = int(table[i].coinc_event_id)
# Remove unneeded rows
table[:] = [
row for row in table
if int(row.coinc_event_id) == target_coinc_event_id
]
target_end_time = table[0].get_end()
coinc_table = table = lsctables.CoincTable.get_table(xmldoc)
table[:] = [
row for row in table
if int(row.coinc_event_id) == target_coinc_event_id
]
table = lsctables.CoincMapTable.get_table(xmldoc)
table[:] = [
row for row in table
if int(row.coinc_event_id) == target_coinc_event_id
]
target_sngl_inspirals = frozenset(row.event_id for row in table)
sngl_inspiral_table = table = lsctables.SnglInspiralTable.get_table(xmldoc)
table[:] = [row for row in table if row.event_id in target_sngl_inspirals]
table = lsctables.ProcessTable.get_table(xmldoc)
table[:] = [row for row in table if row.program == 'gstlal_inspiral']
target_process_ids = frozenset(row.process_id for row in table)
table = lsctables.SearchSummaryTable.get_table(xmldoc)
table[:] = [
row for row in table if target_end_time in row.get_out()
and row.process_id in target_process_ids
]
target_process_ids = frozenset(row.process_id for row in table)
table = lsctables.ProcessTable.get_table(xmldoc)
table[:] = [row for row in table if row.process_id in target_process_ids]
table = lsctables.ProcessParamsTable.get_table(xmldoc)
table[:] = [row for row in table if row.process_id in target_process_ids]
# Shift event times
for row in coinc_inspiral_table:
row.end += delta_t
for row in sngl_inspiral_table:
row.end += delta_t
row.end_time_gmst = lal.GreenwichMeanSiderealTime(row.end)
# The old version of gstlal used to produce the "First Two Years" data set
# stored likelihood in the coinc_event.likelihood column, but newer
# versions store the *natural log* of the likelihood here. The p_astro
# calculation requires this to be log likelihood.
for row in coinc_table:
row.likelihood = np.log(row.likelihood)
# Gstlal stores the template's SVD bank index in the Gamma1 column.
# Fill this in so that we can calculate p_astro
# (see :mod:`gwcelery.tasks.p_astro_gstlal`).
for row in sngl_inspiral_table:
row.Gamma1 = 16
coinc_xml = io.BytesIO()
utils.write_fileobj(xmldoc, coinc_xml)
return coinc_xml.getvalue()