class TestFeatureSchedulerTarget(unittest.TestCase):
def setUp(self) -> None:
self.observatory_model = ObservatoryModel()
self.observatory_model.configure_from_module()
start_time = Time(59853.983, format="mjd", scale="tai")
self.observatory_model.update_state(start_time.unix)
return super().setUp()
def test_constructor(self):
observation = self.make_fbs_observation(note="std")
target = FeatureSchedulerTarget(
observing_script_name="observing_script",
observing_script_is_standard=True,
observation=observation,
)
slew_time, error = self.observatory_model.get_slew_delay(target)
self.assertEqual(error, 0)
self.assertGreater(slew_time, 0.0)
def test_get_script_config(self):
observation = self.make_fbs_observation(note="std")
target = FeatureSchedulerTarget(
observing_script_name="observing_script",
observing_script_is_standard=True,
observation=observation,
)
script_config_expected = {
"targetid":
target.targetid,
"band_filter":
target.filter,
"ra":
Angle(float(observation["RA"][0]),
unit=units.rad).to_string(unit=units.hourangle, sep=":"),
"dec":
Angle(float(observation["dec"][0]),
unit=units.rad).to_string(unit=units.degree, sep=":"),
"name":
observation["note"][0],
"program":
observation["note"][0].rsplit("_", maxsplit=1)[0],
"rot_sky":
target.ang,
"obs_time":
target.obs_time,
"num_exp":
target.num_exp,
"exp_times":
target.exp_times,
"estimated_slew_time":
target.slewtime,
}
script_config_yaml = target.get_script_config()
script_config_unpacked = yaml.safe_load(script_config_yaml)
self.assertEqual(script_config_expected, script_config_unpacked)
def test_get_script_config_cwfs(self):
observation = self.make_fbs_observation(note="cwfs")
target = FeatureSchedulerTarget(
observing_script_name="observing_script",
observing_script_is_standard=True,
observation=observation,
)
script_config_expected = dict(find_target=dict(
az=math.degrees(float(observation["az"][0])),
el=math.degrees(float(observation["alt"][0])),
))
script_config_yaml = target.get_script_config()
script_config_unpacked = yaml.safe_load(script_config_yaml)
self.assertEqual(script_config_expected, script_config_unpacked)
def test_get_script_config_cwfs_with_additional_config(self):
observation = self.make_fbs_observation(note="cwfs")
target = FeatureSchedulerTarget(
observing_script_name="observing_script",
observing_script_is_standard=True,
observation=observation,
script_configuration_cwfs=dict(filter="SDSSg", grating="empty_1"),
)
script_config_expected = dict(
find_target=dict(
az=math.degrees(float(observation["az"][0])),
el=math.degrees(float(observation["alt"][0])),
),
filter="SDSSg",
grating="empty_1",
)
script_config_yaml = target.get_script_config()
script_config_unpacked = yaml.safe_load(script_config_yaml)
self.assertEqual(script_config_expected, script_config_unpacked)
def test_get_script_config_spec(self):
observation = self.make_fbs_observation(note="spec:HD12345")
target = FeatureSchedulerTarget(
observing_script_name="observing_script",
observing_script_is_standard=True,
observation=observation,
)
script_config_expected = {
"object_name":
observation["note"][0],
"object_dec":
Angle(float(observation["dec"][0]),
unit=units.rad).to_string(unit=units.degree, sep=":"),
"object_ra":
Angle(float(observation["RA"][0]),
unit=units.rad).to_string(unit=units.hourangle, sep=":"),
}
script_config_yaml = target.get_script_config()
script_config_unpacked = yaml.safe_load(script_config_yaml)
self.assertEqual(script_config_expected, script_config_unpacked)
def test_get_script_config_spec_with_additional_config(self):
observation = self.make_fbs_observation(note="spec:HD12345")
target = FeatureSchedulerTarget(
observing_script_name="observing_script",
observing_script_is_standard=True,
observation=observation,
script_configuration_spec=dict(
filter_sequence=["SDSSg", "SDSSg"],
grating_sequence=["empty_1", "empty_1"],
),
)
script_config_expected = {
"object_name":
observation["note"][0],
"object_dec":
Angle(float(observation["dec"][0]),
unit=units.rad).to_string(unit=units.degree, sep=":"),
"object_ra":
Angle(float(observation["RA"][0]),
unit=units.rad).to_string(unit=units.hourangle, sep=":"),
"filter_sequence": ["SDSSg", "SDSSg"],
"grating_sequence": ["empty_1", "empty_1"],
}
script_config_yaml = target.get_script_config()
script_config_unpacked = yaml.safe_load(script_config_yaml)
self.assertEqual(script_config_expected, script_config_unpacked)
def test_get_script_config_multiple_observations(self):
filter_obs = "gri"
observations = self.make_fbs_observation("std", filter_obs=filter_obs)
target = FeatureSchedulerTarget(
observing_script_name="observing_script",
observing_script_is_standard=True,
observation=observations,
)
slew_time, error = self.observatory_model.get_slew_delay(target)
script_config = yaml.safe_load(target.get_script_config())
self.assertEqual(len(script_config["exp_times"]), len(filter_obs) * 2)
for filter_name in filter_obs:
self.assertIn(filter_name, script_config["band_filter"])
self.assertEqual(error, 0)
self.assertGreater(slew_time, 0.0)
def make_fbs_observation(self, note, filter_obs="r"):
observations = np.concatenate(
[empty_observation() for _ in range(len(filter_obs))])
ra, dec, _ = self.observatory_model.altaz2radecpa(
self.observatory_model.dateprofile, 65.0, 180.0)
for obs_filter, observation in zip(filter_obs, observations):
observation["RA"] = ra
observation["dec"] = dec
observation["mjd"] = self.observatory_model.dateprofile.mjd
observation["filter"] = obs_filter
observation["exptime"] = 30.0
observation["nexp"] = 2
observation["note"] = note
return observations
class Driver(object):
def __init__(self):
self.log = logging.getLogger("schedulerDriver")
self.params = DriverParameters()
self.location = ObservatoryLocation()
self.observatoryModel = ObservatoryModel(self.location, WORDY)
self.observatoryModel2 = ObservatoryModel(self.location, WORDY)
self.observatoryState = ObservatoryState()
self.sky = AstronomicalSkyModel(self.location)
self.db = FieldsDatabase()
self.build_fields_dict()
self.propid_counter = 0
self.science_proposal_list = []
self.start_time = 0.0
self.time = 0.0
self.targetid = 0
self.survey_started = False
self.isnight = False
self.sunset_timestamp = 0.0
self.sunrise_timestamp = 0.0
self.survey_duration_DAYS = 0.0
self.survey_duration_SECS = self.survey_duration_DAYS * 24 * 60 * 60.0
self.darktime = False
self.mounted_filter = ""
self.unmounted_filter = ""
self.midnight_moonphase = 0.0
self.nulltarget = Target()
self.nulltarget.targetid = -1
self.nulltarget.num_exp = 1
self.nulltarget.exp_times = [0.0]
self.nulltarget.num_props = 1
self.nulltarget.propid_list = [0]
self.nulltarget.need_list = [0.0]
self.nulltarget.bonus_list = [0.0]
self.nulltarget.value_list = [0.0]
self.nulltarget.propboost_list = [1.0]
self.last_winner_target = self.nulltarget.get_copy()
self.deep_drilling_target = None
self.need_filter_swap = False
self.filter_to_unmount = ""
self.filter_to_mount = ""
self.cloud = 0.0
self.seeing = 0.0
def configure_survey(self, survey_conf_file):
prop_conf_path = os.path.dirname(survey_conf_file)
confdict = read_conf_file(survey_conf_file)
self.survey_duration_DAYS = confdict["survey"]["survey_duration"]
self.survey_duration_SECS = self.survey_duration_DAYS * 24 * 60 * 60.0
self.propid_counter = 0
self.science_proposal_list = []
if 'scripted_propconf' in confdict["proposals"]:
scripted_propconflist = confdict["proposals"]["scripted_propconf"]
else:
scripted_propconflist = []
if not isinstance(scripted_propconflist, list):
# turn it into a list with one entry
propconf = scripted_propconflist
scripted_propconflist = []
scripted_propconflist.append(propconf)
self.log.info("configure_survey: scripted proposals %s" % (scripted_propconflist))
for k in range(len(scripted_propconflist)):
self.propid_counter += 1
scripted_prop = ScriptedProposal(self.propid_counter,
os.path.join(prop_conf_path,
"{}".format(scripted_propconflist[k])),
self.sky)
self.science_proposal_list.append(scripted_prop)
if 'areadistribution_propconf' in confdict["proposals"]:
areadistribution_propconflist = confdict["proposals"]["areadistribution_propconf"]
else:
areadistribution_propconflist = []
self.log.info("areadistributionPropConf:%s default" % (areadistribution_propconflist))
if not isinstance(areadistribution_propconflist, list):
# turn it into a list with one entry
propconf = areadistribution_propconflist
areadistribution_propconflist = []
areadistribution_propconflist.append(propconf)
self.log.info("init: areadistribution proposals %s" % (areadistribution_propconflist))
for k in range(len(areadistribution_propconflist)):
self.propid_counter += 1
configfilepath = os.path.join(prop_conf_path, "{}".format(areadistribution_propconflist[k]))
(path, name_ext) = os.path.split(configfilepath)
(name, ext) = os.path.splitext(name_ext)
proposal_confdict = read_conf_file(configfilepath)
self.create_area_proposal(self.propid_counter, name, proposal_confdict)
for prop in self.science_proposal_list:
prop.configure_constraints(self.params)
def configure_duration(self, survey_duration):
self.survey_duration_DAYS = survey_duration
self.survey_duration_SECS = survey_duration * 24 * 60 * 60.0
def configure(self, confdict):
self.params.configure(confdict)
self.log.log(WORDY,
"configure: coadd_values=%s" % (self.params.coadd_values))
self.log.log(WORDY,
"configure: time_balancing=%s" % (self.params.time_balancing))
self.log.log(WORDY,
"configure: timecost_dc=%.3f" % (self.params.timecost_dc))
self.log.log(WORDY,
"configure: timecost_dt=%.3f" % (self.params.timecost_dt))
self.log.log(WORDY,
"configure: timecost_k=%.3f" % (self.params.timecost_k))
self.log.log(WORDY,
"configure: timecost_weight=%.3f" % (self.params.timecost_weight))
self.log.log(WORDY,
"configure: night_boundary=%.1f" % (self.params.night_boundary))
self.log.log(WORDY,
"configure: ignore_sky_brightness=%s" % (self.params.ignore_sky_brightness))
self.log.log(WORDY,
"configure: ignore_airmass=%s" % (self.params.ignore_airmass))
self.log.log(WORDY,
"configure: ignore_clouds=%s" % (self.params.ignore_clouds))
self.log.log(WORDY,
"configure: ignore_seeing=%s" % (self.params.ignore_seeing))
self.log.log(WORDY,
"configure: new_moon_phase_threshold=%.2f" % (self.params.new_moon_phase_threshold))
for prop in self.science_proposal_list:
prop.configure_constraints(self.params)
def configure_location(self, confdict):
self.location.configure(confdict)
self.observatoryModel.location.configure(confdict)
self.observatoryModel2.location.configure(confdict)
self.sky.update_location(self.location)
def configure_observatory(self, confdict):
self.observatoryModel.configure(confdict)
self.observatoryModel2.configure(confdict)
def configure_telescope(self, confdict):
self.observatoryModel.configure_telescope(confdict)
self.observatoryModel2.configure_telescope(confdict)
def configure_rotator(self, confdict):
self.observatoryModel.configure_rotator(confdict)
self.observatoryModel2.configure_rotator(confdict)
def configure_dome(self, confdict):
self.observatoryModel.configure_dome(confdict)
self.observatoryModel2.configure_dome(confdict)
def configure_optics(self, confdict):
self.observatoryModel.configure_optics(confdict)
self.observatoryModel2.configure_optics(confdict)
def configure_camera(self, confdict):
self.observatoryModel.configure_camera(confdict)
self.observatoryModel2.configure_camera(confdict)
def configure_slew(self, confdict):
self.observatoryModel.configure_slew(confdict)
self.observatoryModel2.configure_slew(confdict)
def configure_park(self, confdict):
self.observatoryModel.configure_park(confdict)
self.observatoryModel2.configure_park(confdict)
def create_area_proposal(self, propid, name, config_dict):
self.propid_counter += 1
area_prop = AreaDistributionProposal(propid, name, config_dict, self.sky)
area_prop.configure_constraints(self.params)
self.science_proposal_list.append(area_prop)
def create_sequence_proposal(self, propid, name, config_dict):
self.propid_counter += 1
seq_prop = TimeDistributionProposal(propid, name, config_dict, self.sky)
seq_prop.configure_constraints(self.params)
self.science_proposal_list.append(seq_prop)
def build_fields_dict(self):
sql = "select * from Field"
res = self.db.query(sql)
self.fields_dict = {}
for row in res:
field = Field()
fieldid = row[0]
field.fieldid = fieldid
field.fov_rad = math.radians(row[1])
field.ra_rad = math.radians(row[2])
field.dec_rad = math.radians(row[3])
field.gl_rad = math.radians(row[4])
field.gb_rad = math.radians(row[5])
field.el_rad = math.radians(row[6])
field.eb_rad = math.radians(row[7])
self.fields_dict[fieldid] = field
self.log.log(EXTENSIVE, "buildFieldsTable: %s" % (self.fields_dict[fieldid]))
self.log.info("buildFieldsTable: %d fields" % (len(self.fields_dict)))
def get_fields_dict(self):
return self.fields_dict
def start_survey(self, timestamp, night):
self.start_time = timestamp
self.log.info("start_survey t=%.6f" % timestamp)
self.survey_started = True
for prop in self.science_proposal_list:
prop.start_survey()
self.sky.update(timestamp)
(sunset, sunrise) = self.sky.get_night_boundaries(self.params.night_boundary)
self.log.debug("start_survey sunset=%.6f sunrise=%.6f" % (sunset, sunrise))
# if round(sunset) <= round(timestamp) < round(sunrise):
if sunset <= timestamp < sunrise:
self.start_night(timestamp, night)
self.sunset_timestamp = sunset
self.sunrise_timestamp = sunrise
def end_survey(self):
self.log.info("end_survey")
for prop in self.science_proposal_list:
prop.end_survey()
def start_night(self, timestamp, night):
timeprogress = (timestamp - self.start_time) / self.survey_duration_SECS
self.log.info("start_night t=%.6f, night=%d timeprogress=%.2f%%" %
(timestamp, night, 100 * timeprogress))
self.isnight = True
for prop in self.science_proposal_list:
prop.start_night(timestamp, self.observatoryModel.current_state.mountedfilters, night)
def end_night(self, timestamp, night):
timeprogress = (timestamp - self.start_time) / self.survey_duration_SECS
self.log.info("end_night t=%.6f, night=%d timeprogress=%.2f%%" %
(timestamp, night, 100 * timeprogress))
self.isnight = False
self.last_winner_target = self.nulltarget
self.deep_drilling_target = None
total_filter_visits_dict = {}
total_filter_goal_dict = {}
total_filter_progress_dict = {}
for prop in self.science_proposal_list:
prop.end_night(timestamp)
filter_visits_dict = {}
filter_goal_dict = {}
filter_progress_dict = {}
for filter in self.observatoryModel.filters:
if filter not in total_filter_visits_dict:
total_filter_visits_dict[filter] = 0
total_filter_goal_dict[filter] = 0
filter_visits_dict[filter] = prop.get_filter_visits(filter)
filter_goal_dict[filter] = prop.get_filter_goal(filter)
filter_progress_dict[filter] = prop.get_filter_progress(filter)
total_filter_visits_dict[filter] += filter_visits_dict[filter]
total_filter_goal_dict[filter] += filter_goal_dict[filter]
self.log.debug("end_night propid=%d name=%s filter=%s progress=%.2f%%" %
(prop.propid, prop.name, filter, 100 * filter_progress_dict[filter]))
for filter in self.observatoryModel.filters:
if total_filter_goal_dict[filter] > 0:
total_filter_progress_dict[filter] = \
float(total_filter_visits_dict[filter]) / total_filter_goal_dict[filter]
else:
total_filter_progress_dict[filter] = 0.0
self.log.info("end_night filter=%s progress=%.2f%%" %
(filter, 100 * total_filter_progress_dict[filter]))
previous_midnight_moonphase = self.midnight_moonphase
self.sky.update(timestamp)
(sunset, sunrise) = self.sky.get_night_boundaries(self.params.night_boundary)
self.log.debug("end_night sunset=%.6f sunrise=%.6f" % (sunset, sunrise))
self.sunset_timestamp = sunset
self.sunrise_timestamp = sunrise
next_midnight = (sunset + sunrise) / 2
self.sky.update(next_midnight)
info = self.sky.get_moon_sun_info(numpy.array([0.0]), numpy.array([0.0]))
self.midnight_moonphase = info["moonPhase"]
self.log.info("end_night next moonphase=%.2f%%" % (self.midnight_moonphase))
self.need_filter_swap = False
self.filter_to_mount = ""
self.filter_to_unmount = ""
if self.darktime:
if self.midnight_moonphase > previous_midnight_moonphase:
self.log.info("end_night dark time waxing")
if self.midnight_moonphase > self.params.new_moon_phase_threshold:
self.need_filter_swap = True
self.filter_to_mount = self.unmounted_filter
self.filter_to_unmount = self.mounted_filter
self.darktime = False
else:
self.log.info("end_night dark time waning")
else:
if self.midnight_moonphase < previous_midnight_moonphase:
self.log.info("end_night bright time waning")
if self.midnight_moonphase < self.params.new_moon_phase_threshold:
self.need_filter_swap = True
self.filter_to_mount = self.observatoryModel.params.filter_darktime
max_progress = -1.0
for filter in self.observatoryModel.params.filter_removable_list:
if total_filter_progress_dict[filter] > max_progress:
self.filter_to_unmount = filter
max_progress = total_filter_progress_dict[filter]
self.darktime = True
else:
self.log.info("end_night bright time waxing")
if self.need_filter_swap:
self.log.debug("end_night filter swap %s=>cam=>%s" %
(self.filter_to_mount, self.filter_to_unmount))
def swap_filter(self, filter_to_unmount, filter_to_mount):
self.log.info("swap_filter swap %s=>cam=>%s" % (filter_to_mount, filter_to_unmount))
self.observatoryModel.swap_filter(filter_to_unmount)
self.unmounted_filter = filter_to_unmount
self.mounted_filter = filter_to_mount
return
def update_time(self, timestamp, night):
self.time = timestamp
self.observatoryModel.update_state(self.time)
if not self.survey_started:
self.start_survey(timestamp, night)
if self.isnight:
# if round(timestamp) >= round(self.sunrise_timestamp):
if timestamp >= self.sunrise_timestamp:
self.end_night(timestamp, night)
else:
# if round(timestamp) >= round(self.sunset_timestamp):
if timestamp >= self.sunset_timestamp:
self.start_night(timestamp, night)
return self.isnight
def get_need_filter_swap(self):
return (self.need_filter_swap, self.filter_to_unmount, self.filter_to_mount)
def update_internal_conditions(self, observatory_state, night):
if observatory_state.unmountedfilters != self.observatoryModel.current_state.unmountedfilters:
unmount = observatory_state.unmountedfilters[0]
mount = self.observatoryModel.current_state.unmountedfilters[0]
self.swap_filter(unmount, mount)
for prop in self.science_proposal_list:
prop.start_night(observatory_state.time, observatory_state.mountedfilters, night)
self.time = observatory_state.time
self.observatoryModel.set_state(observatory_state)
self.observatoryState.set(observatory_state)
def update_external_conditions(self, cloud, seeing):
self.cloud = cloud
self.seeing = seeing
return
def select_next_target(self):
if not self.isnight:
return self.nulltarget
targets_dict = {}
ranked_targets_list = []
propboost_dict = {}
sumboost = 0.0
timeprogress = (self.time - self.start_time) / self.survey_duration_SECS
for prop in self.science_proposal_list:
progress = prop.get_progress()
if self.params.time_balancing:
if progress > 0.0:
if timeprogress < 1.0:
needindex = (1.0 - progress) / (1.0 - timeprogress)
else:
needindex = 0.0
if timeprogress > 0.0:
progressindex = progress / timeprogress
else:
progressindex = 1.0
propboost_dict[prop.propid] = needindex / progressindex
else:
propboost_dict[prop.propid] = 1.0
else:
propboost_dict[prop.propid] = 1.0
sumboost += propboost_dict[prop.propid]
if self.deep_drilling_target is not None:
self.log.debug("select_next_target: in deep drilling %s" % str(self.deep_drilling_target))
if self.observatoryModel.is_filter_change_allowed():
constrained_filter = None
else:
constrained_filter = self.observatoryModel.current_state.filter
num_filter_changes = self.observatoryModel.get_number_filter_changes()
delta_burst = self.observatoryModel.get_delta_filter_burst()
delta_avg = self.observatoryModel.get_delta_filter_avg()
self.log.debug("select_next_target: filter changes num=%i tburst=%.1f tavg=%.1f constrained=%s" %
(num_filter_changes, delta_burst, delta_avg, constrained_filter))
for prop in self.science_proposal_list:
propboost_dict[prop.propid] = \
(propboost_dict[prop.propid] * len(self.science_proposal_list) / sumboost) ** self.params.propboost_weight
proptarget_list = prop.suggest_targets(self.time,
self.deep_drilling_target, constrained_filter,
self.cloud, self.seeing)
self.log.log(EXTENSIVE, "select_next_target propid=%d name=%s "
"targets=%d progress=%.2f%% propboost=%.3f" %
(prop.propid, prop.name,
len(proptarget_list), 100 * progress, propboost_dict[prop.propid]))
for target in proptarget_list:
target.num_props = 1
target.propboost = propboost_dict[prop.propid]
target.propid_list = [prop.propid]
target.need_list = [target.need]
target.bonus_list = [target.bonus]
target.value_list = [target.value]
target.propboost_list = [target.propboost]
target.sequenceid_list = [target.sequenceid]
target.subsequencename_list = [target.subsequencename]
target.groupid_list = [target.groupid]
target.groupix_list = [target.groupix]
target.is_deep_drilling_list = [target.is_deep_drilling]
target.is_dd_firstvisit_list = [target.is_dd_firstvisit]
target.remaining_dd_visits_list = [target.remaining_dd_visits]
target.dd_exposures_list = [target.dd_exposures]
target.dd_filterchanges_list = [target.dd_filterchanges]
target.dd_exptime_list = [target.dd_exptime]
fieldfilter = (target.fieldid, target.filter)
if fieldfilter in targets_dict:
if self.params.coadd_values:
targets_dict[fieldfilter][0] = targets_dict[fieldfilter][0].get_copy()
targets_dict[fieldfilter][0].need += target.need
targets_dict[fieldfilter][0].bonus += target.bonus
targets_dict[fieldfilter][0].value += target.value
targets_dict[fieldfilter][0].propboost += target.propboost
if target.is_deep_drilling:
# overrides to make the coadded target a consistent deep drilling
targets_dict[fieldfilter][0].is_deep_drilling = target.is_deep_drilling
targets_dict[fieldfilter][0].is_dd_firstvisit = target.is_dd_firstvisit
targets_dict[fieldfilter][0].remaining_dd_visits = target.remaining_dd_visits
targets_dict[fieldfilter][0].dd_exposures = target.dd_exposures
targets_dict[fieldfilter][0].dd_filterchanges = target.dd_filterchanges
targets_dict[fieldfilter][0].dd_exptime = target.dd_exptime
targets_dict[fieldfilter][0].sequenceid = target.sequenceid
targets_dict[fieldfilter][0].subsequencename = target.subsequencename
targets_dict[fieldfilter][0].groupid = target.groupid
targets_dict[fieldfilter][0].groupix = target.groupix
else:
# new target to coadd is not deep drilling
if not targets_dict[fieldfilter][0].is_deep_drilling:
# coadded target is not deep drilling
# overrides with new sequence information
targets_dict[fieldfilter][0].sequenceid = target.sequenceid
targets_dict[fieldfilter][0].subsequencename = target.subsequencename
targets_dict[fieldfilter][0].groupid = target.groupid
targets_dict[fieldfilter][0].groupix = target.groupix
# if coadded target is already deep drilling, don't override
targets_dict[fieldfilter][0].num_props += 1
targets_dict[fieldfilter][0].propid_list.append(prop.propid)
targets_dict[fieldfilter][0].need_list.append(target.need)
targets_dict[fieldfilter][0].bonus_list.append(target.bonus)
targets_dict[fieldfilter][0].value_list.append(target.value)
targets_dict[fieldfilter][0].propboost_list.append(target.propboost)
targets_dict[fieldfilter][0].sequenceid_list.append(target.sequenceid)
targets_dict[fieldfilter][0].subsequencename_list.append(target.subsequencename)
targets_dict[fieldfilter][0].groupid_list.append(target.groupid)
targets_dict[fieldfilter][0].groupix_list.append(target.groupix)
targets_dict[fieldfilter][0].is_deep_drilling_list.append(target.is_deep_drilling)
targets_dict[fieldfilter][0].is_dd_firstvisit_list.append(target.is_dd_firstvisit)
targets_dict[fieldfilter][0].remaining_dd_visits_list.append(
target.remaining_dd_visits)
targets_dict[fieldfilter][0].dd_exposures_list.append(target.dd_exposures)
targets_dict[fieldfilter][0].dd_filterchanges_list.append(target.dd_filterchanges)
targets_dict[fieldfilter][0].dd_exptime_list.append(target.dd_exptime)
else:
targets_dict[fieldfilter].append(target)
else:
targets_dict[fieldfilter] = [target]
filtercost = self.compute_filterchange_cost() * self.params.filtercost_weight
for fieldfilter in targets_dict:
slewtime = self.observatoryModel.get_slew_delay(targets_dict[fieldfilter][0])
if slewtime >= 0:
timecost = self.compute_slewtime_cost(slewtime) * self.params.timecost_weight
for target in targets_dict[fieldfilter]:
target.slewtime = slewtime
if target.filter != self.observatoryModel.current_state.filter:
target.cost = timecost + filtercost
else:
target.cost = timecost
target.rank = (target.value * target.propboost) - target.cost
ranked_targets_list.append((-target.rank, target))
sorted_list = sorted(ranked_targets_list, key=itemgetter(0))
winner_found = False
while len(sorted_list) > 0 and not winner_found:
winner_target = sorted_list.pop(0)[1]
self.observatoryModel2.set_state(self.observatoryState)
self.observatoryModel2.observe(winner_target)
if winner_target.is_dd_firstvisit:
ttime = self.observatoryModel2.get_deep_drilling_time(winner_target)
else:
ttime = 0.0
ntime = self.observatoryModel2.current_state.time + ttime + 30.0
if ntime < self.sunrise_timestamp:
self.observatoryModel2.update_state(ntime)
if self.observatoryModel2.current_state.tracking:
self.targetid += 1
winner_target.targetid = self.targetid
winner_target.time = self.time
winner_found = True
else:
self.log.debug("select_next_target: target rejected ttime=%.1f %s" %
(ttime, str(winner_target)))
self.log.debug("select_next_target: state rejected %s" %
str(self.observatoryModel2.current_state))
else:
self.log.debug("select_next_target: target rejected ttime=%.1f %s" %
(ttime, str(winner_target)))
self.log.debug("select_next_target: rejected due to end of night")
if ttime == 0.0:
# ttime == 0 means it is a regular visit (not DD)
# if there is no time left for a single visit then quit
break
if winner_found:
if not self.params.coadd_values:
first_target = targets_dict[(winner_target.fieldid, winner_target.filter)][0]
if first_target.propid != winner_target.propid:
winner_target.copy_driver_state(first_target)
self.last_winner_target = winner_target.get_copy()
else:
self.last_winner_target = self.nulltarget
return self.last_winner_target
def register_observation(self, observation):
target_list = []
if observation.targetid > 0:
if self.observation_fulfills_target(observation, self.last_winner_target):
observation = self.last_winner_target
else:
self.log.info("register_observation: unexpected observation %s" % str(observation))
for prop in self.science_proposal_list:
target = prop.register_observation(observation)
if target is not None:
target_list.append(target)
self.last_observation = observation.get_copy()
if self.last_observation.is_deep_drilling and (self.last_observation.remaining_dd_visits > 1):
self.deep_drilling_target = self.last_observation
else:
self.deep_drilling_target = None
return target_list
def compute_slewtime_cost(self, slewtime):
cost = (self.params.timecost_k / (slewtime + self.params.timecost_dt) -
self.params.timecost_dc - self.params.timecost_cref) / (1.0 - self.params.timecost_cref)
#cost = self.params.timecost_k / (slewtime + self.params.timecost_dt) - self.params.timecost_dc
return cost
def compute_filterchange_cost(self):
t = self.observatoryModel.get_delta_last_filterchange()
T = self.observatoryModel.params.filter_max_changes_avg_interval
if t < T:
cost = 1.0 - t / T
else:
cost = 0.0
return cost
def observation_fulfills_target(self, observ, target):
return (observ.fieldid == target.fieldid) and (observ.filter == target.filter)