def test_pbar_hide_output(self):
"""Empty output when not showing progressbar"""
pb = utils.pbar(self.iterable, show=False, fd=self.output)
self.assertEqual(list(pb), self.iterable)
self.assertEqual(self.output.getvalue(), '')
pb = utils.pbar(self.iterable, show=True, fd=self.output)
self.assertEqual(list(pb), self.iterable)
self.assertNotEqual(self.output.getvalue(), '')
def update_additional_local_tsv():
"""Get location tsv data for all stations"""
station_numbers = Network().station_numbers()
for type in ['eventtime', 'detector_timing_offsets']:
try:
mkdir(path.join(LOCAL_BASE, type))
except OSError:
pass
for number in pbar(station_numbers):
url = API.src_urls[type].format(station_number=number,
year='', month='', day='', hour='')
try:
data = API._retrieve_url(url.strip('/'), base=SRC_BASE)
except:
print 'Failed to get %s data for station %d' % (type, number)
continue
data = '\n'.join(d for d in data.split('\n')
if len(d) and d[0] != '#')
if data:
tsv_path = path.join(LOCAL_BASE,
url.strip('/') + extsep + 'tsv')
with open(tsv_path, 'w') as tsvfile:
tsvfile.write(data)
type = 'station_timing_offsets'
network = HiSPARCNetwork()
try:
mkdir(path.join(LOCAL_BASE, type))
except OSError:
pass
for number1, number2 in pbar(combinations(station_numbers, 2)):
distance = network.calc_distance_between_stations(number1, number2)
if distance is None or distance > 1e3:
continue
try:
mkdir(path.join(LOCAL_BASE, type, str(number1)))
except OSError:
pass
url = API.src_urls[type].format(station_1=number1, station_2=number2)
try:
data = API._retrieve_url(url.strip('/'), base=SRC_BASE)
except:
print ('Failed to get %s data for station pair %d-%d' %
(type, number1, number2))
continue
data = '\n'.join(d for d in data.split('\n') if len(d) and d[0] != '#')
if data:
tsv_path = path.join(LOCAL_BASE, url.strip('/') + extsep + 'tsv')
with open(tsv_path, 'w') as tsvfile:
tsvfile.write(data)
def store_coincidences(self):
print "Storing coincidences..."
if '/coincidences' not in self.data:
group = self.data.create_group('/', 'coincidences')
group._v_attrs.cluster = self.cluster
self.c_index = []
self.coincidences = self.data.create_table(group, 'coincidences',
storage.Coincidence)
self.observables = self.data.create_table(group, 'observables',
storage.EventObservables)
for coincidence in pbar(self.data.root.c_index):
self.store_coincidence(coincidence)
c_index = self.data.create_vlarray(group, 'c_index',
tables.UInt32Col())
for coincidence in self.c_index:
c_index.append(coincidence)
c_index.flush()
self.c_index = c_index
else:
# Force new cluster geometry
group = self.data.get_node('/', 'coincidences')
group._v_attrs.cluster = self.cluster
def write_jsons(data):
if not os.path.exists(os.path.join(EVENT_DISPLAY_DIR, 'data/')):
os.mkdir(os.path.join(EVENT_DISPLAY_DIR, 'data'))
# Create subcluster station location and coincidences JSONs
for subcluster in Network().subclusters():
subcluster_name = subcluster['name'].lower().replace(' ', '_')
stations = build_station_json(data, subcluster['number'])
with open('data/stations_' + subcluster_name + '.json', 'w') as f:
json.dump(stations, f)
coincidences = build_coincidence_json(data, subcluster_name)
with open('data/coincidences_' + subcluster_name + '.json', 'w') as f:
json.dump(coincidences, f)
# Create network station location and coincidences JSONs
stations = build_station_json(data)
with open('data/stations_network.json', 'w') as f:
json.dump(stations, f)
coincidences = build_coincidence_json(data)
with open('data/coincidences_network.json', 'w') as f:
json.dump(coincidences, f)
# Create station events JSONs
for station in pbar(STATIONS):
events = build_events_json(data, station)
with open('data/events_s%d.json' % station, 'w') as f:
json.dump(events, f)
def write_jsons(data):
if not os.path.exists(os.path.join(EVENT_DISPLAY_DIR, 'data/')):
os.mkdir(os.path.join(EVENT_DISPLAY_DIR, 'data'))
# Create subcluster station location and coincidences JSONs
for subcluster in Network().subclusters():
subcluster_name = subcluster['name'].lower().replace(' ', '_')
stations = build_station_json(data, subcluster['number'])
with open('data/stations_' + subcluster_name + '.json', 'w') as f:
json.dump(stations, f)
coincidences = build_coincidence_json(data, subcluster_name)
with open('data/coincidences_' + subcluster_name + '.json', 'w') as f:
json.dump(coincidences, f)
# Create network station location and coincidences JSONs
stations = build_station_json(data)
with open('data/stations_network.json', 'w') as f:
json.dump(stations, f)
coincidences = build_coincidence_json(data)
with open('data/coincidences_network.json', 'w') as f:
json.dump(coincidences, f)
# Create station events JSONs
for station in pbar(STATIONS):
events = build_events_json(data, station)
with open('data/events_s%d.json' % station, 'w') as f:
json.dump(events, f)
def store_coincidences(self):
print "Storing coincidences..."
if '/coincidences' not in self.data:
group = self.data.create_group('/', 'coincidences')
group._v_attrs.cluster = self.cluster
self.c_index = []
self.coincidences = self.data.create_table(group,
'coincidences',
storage.Coincidence)
self.observables = self.data.create_table(group, 'observables',
storage.EventObservables)
for coincidence in pbar(self.data.root.c_index):
self.store_coincidence(coincidence)
c_index = self.data.create_vlarray(group, 'c_index',
tables.UInt32Col())
for coincidence in self.c_index:
c_index.append(coincidence)
c_index.flush()
self.c_index = c_index
else:
# Force new cluster geometry
group = self.data.get_node('/', 'coincidences')
group._v_attrs.cluster = self.cluster
def write_jsons(data):
if not os.path.exists('data/'):
os.mkdir('data')
for subcluster in Network().subclusters():
subcluster_name = subcluster['name'].lower().replace(' ', '_')
stations = build_station_json(data, subcluster['number'])
with open('data/stations_' + subcluster_name + '.json', 'w') as f:
json.dump(stations, f)
coincidences = build_coincidence_json(data, subcluster_name)
with open('data/coincidences_' + subcluster_name + '.json', 'w') as f:
json.dump(coincidences, f)
stations = build_station_json(data)
with open('data/stations_network.json', 'w') as f:
json.dump(stations, f)
coincidences = build_coincidence_json(data)
with open('data/coincidences_network.json', 'w') as f:
json.dump(coincidences, f)
for station in pbar(STATIONS):
events = build_events_json(data, station)
with open('data/events_s%d.json' % station, 'w') as f:
json.dump(events, f)
def migrate():
"""
Find unmigrated tables in datastore
migrate tables
check datastore again for unmigrated tables
"""
logging.info('******************')
logging.info('Starting migration')
logging.info('******************')
queue = get_queue(DATASTORE_PATH)
print('migrating:')
for path in pbar(queue.keys()):
logging.info('Migrating: %s' % path)
with tables.open_file(path, 'a') as data:
migration = MigrateSingles(data)
for table in queue[path]:
logging.debug('Processing table: %s' % table)
migration.migrate_table(table)
queue = get_queue(DATASTORE_PATH)
if queue:
logging.error('Found unprocessed tables after migration')
for path in queue.keys():
logging.error('Unprocessed tables in: %s' % path)
for table in queue[path]:
logging.error('%s' % table)
else:
logging.info('********************')
logging.info('Migration succesful!')
logging.info('********************')
def get_queue(datastore_path):
queue = {}
logging.info('Searching for unmigrated singles tables')
print('Looking for singles tables in datastore.')
# Singles tables were added in Feb, 2016.
for fn in pbar(glob.glob(datastore_path + '/201[6,7]/*/*h5')):
singles_tables = []
with tables.open_file(fn, 'r') as data:
for node in data.walk_nodes('/', 'Table'):
table_path = node._v_pathname
if '/singles' in table_path:
if 'singles_old' in table_path:
continue
type_ = type(node.description.mas_ch1_low)
if type_ == tables.UInt16Col:
logging.debug('Found: %s' % table_path)
singles_tables.append(table_path)
elif type_ == tables.Int32Col:
logging.debug('Skipping migrated: %s' % table_path)
continue
else:
logging.error('%s in unknown format!' % table_path)
if singles_tables:
queue[fn] = singles_tables
logging.info('Found %d tables in %s' % (len(singles_tables), fn))
n = sum(len(v) for v in queue.itervalues())
logging.info('Found %d unmigrated tables '
'in %d datastore files.' % (n, len(queue)))
return queue
def make_datasets_failed_reconstructions_scatter(data):
global dt1, dt2, phis_sim, phis_rec
global gdt1, gdt2, gphis_sim, gphis_rec
group = data.root.simulations.E_1PeV.zenith_22_5.shower_0
observables = group.observables
coincidences = group.coincidences
dt1, dt2, phis_sim, phis_rec = [], [], [], []
gdt1, gdt2, gphis_sim, gphis_rec = [], [], [], []
for event, coincidence in pbar(izip(observables, coincidences),
len(observables)):
assert event['id'] == coincidence['id']
if min(event['n1'], event['n3'], event['n4']) >= 1.:
theta, phi = reconstruct_angle(event, 10.)
assert not isnan(phi)
if isnan(theta):
dt1.append(event['t1'] - event['t3'])
dt2.append(event['t1'] - event['t4'])
phis_sim.append(coincidence['phi'])
phis_rec.append(phi)
else:
gdt1.append(event['t1'] - event['t3'])
gdt2.append(event['t1'] - event['t4'])
gphis_sim.append(coincidence['phi'])
gphis_rec.append(phi)
def perform_simulations():
cq = CorsikaQuery(OVERVIEW)
s = set()
for energy in pbar(cq.available_parameters('energy', particle='proton')):
for zenith in cq.available_parameters('zenith',
particle='proton',
energy=energy):
sims = cq.simulations(particle='proton',
zenith=zenith,
energy=energy,
iterator=False)
selected_seeds = cq.seeds(sims)
n = min(len(selected_seeds), N)
for seeds in choice(selected_seeds, n, replace=False):
if seeds in s:
continue
if not os.path.exists(
'/data/hisparc/corsika/data/{seeds}/corsika.h5'.format(
seeds=seeds)):
continue
s.add(seeds)
if energy >= 16:
perform_job(seeds, 'long')
elif energy >= 14.5:
perform_job(seeds, 'generic')
else:
perform_job(seeds, 'short')
cq.finish()
def binned_stat_idx(events, idx_ranges):
stats = {}
for field_name in FIELD_NAMES:
if field_name == 'event_rate':
stats[field_name] = get_event_rate(idx_ranges)
else:
stats[field_name] = []
for start, stop in pbar(zip(idx_ranges[:-1], idx_ranges[1:])):
slice = events.read(start, stop)
for field, field_name in zip(FIELDS, FIELD_NAMES):
if field_name == 'event_rate':
# Event time determined above
continue
elif field_name == 'mpv':
# Simply the MPV value, not fraction of bad MPV
mpvs = reconstruct_mpv(slice)
stats[field_name].append(mpvs)
elif isinstance(field, tuple):
data = column_stack(slice[f] for f in field)
stats[field_name].append(frac_bad(data))
else:
data = slice[field]
stats[field_name].append(frac_bad(data))
for field_name in FIELD_NAMES:
stats[field_name] = array(stats[field_name]).T
return stats
def make_datasets_failed_reconstructions_scatter(data):
global dt1, dt2, phis_sim, phis_rec
global gdt1, gdt2, gphis_sim, gphis_rec
group = data.root.simulations.E_1PeV.zenith_22_5.shower_0
observables = group.observables
coincidences = group.coincidences
dt1, dt2, phis_sim, phis_rec = [], [], [], []
gdt1, gdt2, gphis_sim, gphis_rec = [], [], [], []
for event, coincidence in pbar(izip(observables, coincidences),
len(observables)):
assert event['id'] == coincidence['id']
if min(event['n1'], event['n3'], event['n4']) >= 1.:
theta, phi = reconstruct_angle(event, 10.)
assert not isnan(phi)
if isnan(theta):
dt1.append(event['t1'] - event['t3'])
dt2.append(event['t1'] - event['t4'])
phis_sim.append(coincidence['phi'])
phis_rec.append(phi)
else:
gdt1.append(event['t1'] - event['t3'])
gdt2.append(event['t1'] - event['t4'])
gphis_sim.append(coincidence['phi'])
gphis_rec.append(phi)
def test_pbar_generator(self):
"""Return original generator, not a progressbar"""
generator = (x for x in self.iterable)
pb = utils.pbar(generator)
self.assertIsInstance(pb, types.GeneratorType)
self.assertEqual(list(pb), self.iterable)
def test_pbar_generator_known_length(self):
"""Return progressbar for generator with known length"""
generator = (y for y in self.iterable)
pb = utils.pbar(generator, length=len(self.iterable), fd=self.output)
self.assertIsInstance(pb, progressbar.ProgressBar)
self.assertEqual(list(pb), self.iterable)
def foreignkey_to_m2m(apps, schema_editor):
"""Backwards migrations"""
fk_model = apps.get_model('coincidences', 'Event')
if fk_model.objects.all().count():
print('')
for event in pbar(fk_model.objects.all().iterator(), length=fk_model.objects.all().count()):
event.coincidence.events.add(event)
def sort(self):
"""Sort the table"""
chunk = self.nrows_in_chunk
nrows = self.nrows
parts = int(nrows / chunk) + 1
if parts == 1:
if self.progress:
print("Sorting table in memory and writing to disk.")
self._sort_chunk(self.outtable, 0, nrows)
else:
if self.progress:
print("Sorting in %d chunks of %d rows:" % (parts, chunk))
for idx, start in pbar(enumerate(range(0, nrows, chunk)),
length=parts,
show=self.progress):
table_name = 'temp_table_%d' % idx
table = self.hdf5_temp.create_table('/',
table_name,
self.description,
expectedrows=chunk)
iterator = self._sort_chunk(table, start, start + chunk)
self._iterators.append(iterator)
rowbuf = self.outtable._get_container(self._BUFSIZE)
idx = 0
if self.progress:
print("Merging:")
for keyedrow in pbar(merge(*self._iterators),
length=nrows,
show=self.progress):
x, row = keyedrow
if idx == self._BUFSIZE:
self.outtable.append(rowbuf)
self.outtable.flush()
idx = 0
rowbuf[idx] = row.fetch_all_fields()
idx += 1
# store last lines in buffer
self.outtable.append(rowbuf[0:idx])
self.outtable.flush()
def download_events_data(data):
"""Download event data for each station into a separate table"""
for station in pbar(STATIONS):
group = '/s%d' % station
if group not in data:
download_data(data, group, station, start=START, end=END,
progress=False)
def get_combined_results():
zenith = []
zenith_in = []
azimuth = []
azimuth_in = []
energy_in = []
size_in = []
r_in = []
zenith_init = []
azimuth_init = []
energy_init = []
size_init = []
r_init = []
for path in pbar(glob.glob(PATHS)):
with tables.open_file(path, 'r') as data:
recs = data.root.coincidences.reconstructions
filtered_recs = recs.read_where(
's501 & s502 & s503 & s504 & s505 & s506')
zenith.extend(degrees(filtered_recs['zenith']))
zenith_in.extend(degrees(filtered_recs['reference_zenith']))
azimuth.extend(degrees(filtered_recs['azimuth']))
azimuth_in.extend(degrees(filtered_recs['reference_azimuth']))
energy_in.extend(log10(filtered_recs['reference_energy']))
size_in.extend(log10(filtered_recs['reference_size']))
r_in.extend(
sqrt(filtered_recs['reference_x']**2 +
filtered_recs['reference_y']**2))
zenith_init.extend(degrees(recs.col('reference_zenith')))
azimuth_init.extend(degrees(recs.col('reference_azimuth')))
energy_init.extend(log10(recs.col('reference_energy')))
size_init.extend(log10(recs.col('reference_size')))
r_init.extend(
sqrt(recs.col('reference_x')**2 + recs.col('reference_y')**2))
zenith = array(zenith)
filter = ~isnan(zenith)
zenith = zenith.compress(filter)
zenith_in = array(zenith_in).compress(filter)
azimuth = array(azimuth).compress(filter)
azimuth_in = array(azimuth_in).compress(filter)
energy_in = array(energy_in).compress(filter)
size_in = array(size_in).compress(filter)
r_in = array(r_in).compress(filter)
zenith_init = array(zenith_init)
azimuth_init = array(azimuth_init)
energy_init = array(energy_init)
size_init = array(size_init)
r_init = array(r_init)
print sum(filter), len(filter)
return (zenith, zenith_in, azimuth, azimuth_in, energy_in, size_in, r_in,
zenith_init, azimuth_init, energy_init, size_init, r_init)
def get_pair_distance_energy_array(distances, energies, n=8):
results = []
for distance in pbar(distances):
sens = DistancePairAreaEnergySensitivity(distance=distance, n=n)
sens.main()
areas = sens.get_area_energy(energies)
results.append(areas)
results = np.array(results)
return results
def update_local_json():
for type in pbar(['stations', 'subclusters', 'clusters', 'countries']):
update_toplevel_json(type)
for arg_type, type in [('stations', 'station_info'),
('subclusters', 'stations_in_subcluster'),
('clusters', 'subclusters_in_cluster'),
('countries', 'clusters_in_country')]:
update_sublevel_json(arg_type, type)
def sort(self):
"""Sort the table"""
chunk = self.nrows_in_chunk
nrows = self.nrows
parts = int(nrows / chunk) + 1
if parts == 1:
if self.progress:
print("Sorting table in memory and writing to disk.")
self._sort_chunk(self.outtable, 0, nrows)
else:
if self.progress:
print("Sorting in %d chunks of %d rows:" % (parts, chunk))
for idx, start in pbar(enumerate(range(0, nrows, chunk)),
length=parts, show=self.progress):
table_name = 'temp_table_%d' % idx
table = self.hdf5_temp.create_table('/', table_name,
self.description,
expectedrows=chunk)
iterator = self._sort_chunk(table, start, start + chunk)
self._iterators.append(iterator)
rowbuf = self.outtable._get_container(self._BUFSIZE)
idx = 0
if self.progress:
print("Merging:")
for keyedrow in pbar(merge(*self._iterators), length=nrows,
show=self.progress):
x, row = keyedrow
if idx == self._BUFSIZE:
self.outtable.append(rowbuf)
self.outtable.flush()
idx = 0
rowbuf[idx] = row.fetch_all_fields()
idx += 1
# store last lines in buffer
self.outtable.append(rowbuf[0:idx])
self.outtable.flush()
def m2m_to_foreignkey(apps, schema_editor):
"""Forwards migrations"""
m2m_model = apps.get_model('coincidences', 'Coincidence')
coincidences = m2m_model.objects.all().annotate(n_events=Count('events')).exclude(n_events=0)
if coincidences.count():
print('')
for coincidence in pbar(coincidences.iterator(), length=coincidences.count()):
for event in coincidence.events.all():
event.coinc = coincidence
event.save()
def serialiseddatafield_to_arrayfield(apps, schema_editor):
"""Forwards migrations"""
model = apps.get_model('histograms', 'NetworkHistogram')
if model.objects.all().count():
print('')
for histogram in pbar(model.objects.all().iterator(),
length=model.objects.all().count()):
histogram.bins = histogram.old_bins
histogram.values = histogram.old_values
histogram.save()
def download_events_data(data):
"""Download event data for each station into a separate table"""
for station in pbar(STATIONS):
group = '/s%d' % station
if group not in data:
download_data(data,
group,
station,
start=START,
end=END,
progress=False)
def serialiseddatafield_to_arrayfield(apps, schema_editor):
"""Forwards migrations"""
model = apps.get_model('coincidences', 'Event')
if model.objects.all().count():
print('')
for event in pbar(model.objects.all().iterator(),
length=model.objects.all().count()):
event.pulseheights = mv_to_adc(event.old_pulseheights)
event.integrals = mvns_to_adcsample(event.old_integrals)
event.traces = traces_mv_to_adc(event.old_traces)
if not (len(event.traces[0]) == len(event.traces[-1])):
event.traces = event.traces[:2]
event.save()
def generate_json():
"""Get the API info data for each station"""
station_numbers = Network().station_numbers()
station_info = {}
for number in pbar(station_numbers):
try:
station = Station(number)
station_info[number] = station.info
except:
continue
return station_info
def reconstruct_angles(self, coincidences_group):
coincidences_group = self.data.get_node(coincidences_group)
self.data_group = coincidences_group
coincidences = coincidences_group.coincidences
self.cluster = coincidences_group._v_attrs.cluster
self.results_group._v_attrs.cluster = self.cluster
sel_coincidences = coincidences.read_where('N >= 1')
for coincidence in pbar(sel_coincidences):
self.reconstruct_individual_stations(coincidence)
self.reconstruct_cluster_stations(coincidence)
self.data.flush()
def generate_json():
"""Get the API info data for each station"""
station_numbers = Network().station_numbers()
station_info = {}
for number in pbar(station_numbers):
try:
station = Station(number)
station_info[number] = station.info
except:
continue
return station_info
def reconstruct_angles(self, coincidences_group):
coincidences_group = self.data.get_node(coincidences_group)
self.data_group = coincidences_group
coincidences = coincidences_group.coincidences
self.cluster = coincidences_group._v_attrs.cluster
self.results_group._v_attrs.cluster = self.cluster
sel_coincidences = coincidences.read_where('N >= 3')
for coincidence in pbar(sel_coincidences):
self.reconstruct_individual_stations(coincidence)
self.reconstruct_cluster_stations(coincidence)
self.data.flush()
def multiple_jobs(n, energy, particle, zenith, azimuth, queue, corsika):
"""Use this to sumbit multiple jobs to Stoomboot
:param n: Number of jobs to submit
:param energy: log10(E[eV]) energy of primary particle
:param particle: Particle kind (as string, see
:mod:`~sapphire.corsika.particles` for possibilities)
:param zenith: Zenith angle in degrees of the primary particle
:param azimuth: Azimuth angle in degrees of the primary particle
:param queue: Stoomboot queue to submit to
:param corsika: Name of the CORSIKA executable to use
"""
print textwrap.dedent("""\
Batch submitting jobs to Stoomboot:
Number of jobs {n}
Particle energy 10^{e} eV
Primary particle {p}
Zenith angle {z} degrees
Azimuth angle {a} degrees
Stoomboot queue {q}
CORSIKA executable {c}
""".format(n=n,
e=energy,
p=particle,
z=zenith,
a=azimuth,
q=queue,
c=corsika))
available_slots = check_queue(queue)
if available_slots <= 0:
n = 0
print 'Submitting no jobs because queue is full.'
return
elif available_slots < n:
n = available_slots
print 'Submitting {n} jobs because queue is almost full.'.format(n=n)
for _ in pbar(xrange(n)):
batch = CorsikaBatch(energy=energy,
particle=particle,
zenith=zenith,
azimuth=azimuth,
queue=queue,
corsika=corsika)
batch.run()
print 'Done.'
def serialiseddatafield_to_arrayfield(apps, schema_editor):
"""Forwards migrations"""
model = apps.get_model('histograms', 'DailyDataset')
multi_model = apps.get_model('histograms', 'MultiDailyDataset')
if model.objects.all().count():
print('')
for dataset in pbar(model.objects.all().iterator(),
length=model.objects.all().count()):
if dataset.type.slug in ['barometer', 'temperature']:
dataset.x = dataset.old_x
dataset.y = dataset.old_y
dataset.save()
else:
multi_model.objects.create(source=dataset.source,
type=dataset.type,
x=dataset.old_x,
y=dataset.old_y)
dataset.delete()
def get_tcc_values(data, force_new=False):
if 'tcc' in data.root and not force_new:
return data.root.tcc.read()
else:
h_events = data.root.hisparc.cluster_kascade.station_601.events
c_index = data.root.kascade.c_index
tcc = []
for idx in pbar(c_index):
event = h_events[idx['h_idx']]
value = calculate_tcc(event)
tcc.append(value)
tcc = array(tcc)
if 'tcc' in data.root:
data.remove_node('/tcc')
data.create_array('/', 'tcc', tcc)
return tcc
def get_integrals(data):
blobs = data.get_node(GROUP, 'blobs')
events = data.get_node(GROUP, 'events')
raw_integrals = []
filtered_integrals = []
source_integrals = []
for event in pbar(events[:2000]):
# traces = get_traces_from_api(API_STATION, event)
traces = get_traces_from_blobs(event, blobs)
raw, filtered = determine_integrals(traces, event)
source = event['integrals']
raw_integrals.append(raw)
filtered_integrals.append(filtered)
source_integrals.append(source)
return (array(raw_integrals), array(filtered_integrals),
array(source_integrals))
def serialiseddatafield_to_arrayfield(apps, schema_editor):
"""Forwards migrations"""
model = apps.get_model('histograms', 'DailyDataset')
multi_model = apps.get_model('histograms', 'MultiDailyDataset')
if model.objects.all().count():
print('')
for dataset in pbar(model.objects.all().iterator(), length=model.objects.all().count()):
if dataset.type.slug in ['barometer', 'temperature']:
dataset.x = dataset.old_x
dataset.y = dataset.old_y
dataset.save()
else:
multi_model.objects.create(
source=dataset.source,
type=dataset.type,
x=dataset.old_x,
y=dataset.old_y)
dataset.delete()
def get_tcc_values(data, force_new=False):
if 'tcc' in data.root and not force_new:
return data.root.tcc.read()
else:
h_events = data.root.hisparc.cluster_kascade.station_601.events
c_index = data.root.kascade.c_index
tcc = []
for idx in pbar(c_index):
event = h_events[idx['h_idx']]
value = calculate_tcc(event)
tcc.append(value)
tcc = array(tcc)
if 'tcc' in data.root:
data.remove_node('/tcc')
data.create_array('/', 'tcc', tcc)
return tcc
def reconstruct_core_positions(self, hisparc_group, kascade_group, tcc):
hisparc_group = self.data.get_node(hisparc_group)
hisparc_table = self.data.get_node(hisparc_group, 'events')
c_index = self.data.get_node(kascade_group, 'c_index')
kascade_table = self.data.get_node(kascade_group, 'events')
self.cluster = hisparc_group._v_attrs.cluster
self._store_cluster_with_results()
for idx, tcc_value in pbar(zip(c_index[:self.N], tcc)):
hisparc_event = hisparc_table[idx['h_idx']]
kascade_event = kascade_table[idx['k_idx']]
if tcc_value >= 10:
x, y, N = self.reconstruct_core_position(hisparc_event)
self.store_reconstructed_event(hisparc_event,
kascade_event, x, y, N)
self.results_table.flush()
def reconstruct_core_positions(self, hisparc_group, kascade_group, tcc):
hisparc_group = self.data.get_node(hisparc_group)
hisparc_table = self.data.get_node(hisparc_group, 'events')
c_index = self.data.get_node(kascade_group, 'c_index')
kascade_table = self.data.get_node(kascade_group, 'events')
self.cluster = hisparc_group._v_attrs.cluster
self._store_cluster_with_results()
for idx, tcc_value in pbar(zip(c_index[:self.N], tcc)):
hisparc_event = hisparc_table[idx['h_idx']]
kascade_event = kascade_table[idx['k_idx']]
if tcc_value >= 10:
x, y, N = self.reconstruct_core_position(hisparc_event)
self.store_reconstructed_event(hisparc_event, kascade_event, x,
y, N)
self.results_table.flush()
def update_local_tsv():
"""Get configuration tsv data for all stations"""
station_numbers = Network().station_numbers()
for type in ['gps', 'trigger', 'layout', 'voltage', 'current',
'electronics']:
subdir = API.src_urls[type].split('/')[0]
try:
mkdir(path.join(LOCAL_BASE, subdir))
except OSError:
pass
for number in pbar(station_numbers):
url = API.src_urls[type].format(station_number=number)
try:
get_and_store_tsv(url)
except:
if type != 'layout':
print 'Failed to get %s for station %d' % (type, number)
continue
def multiple_jobs(n, energy, particle, zenith, azimuth, queue, corsika):
"""Use this to sumbit multiple jobs to Stoomboot
:param n: Number of jobs to submit
:param energy: log10(E[eV]) energy of primary particle
:param particle: Particle kind (as string, see
:mod:`~sapphire.corsika.particles` for possibilities)
:param zenith: Zenith angle in degrees of the primary particle
:param azimuth: Azimuth angle in degrees of the primary particle
:param queue: Stoomboot queue to submit to
:param corsika: Name of the CORSIKA executable to use
"""
print textwrap.dedent("""\
Batch submitting jobs to Stoomboot:
Number of jobs {n}
Particle energy 10^{e} eV
Primary particle {p}
Zenith angle {z} degrees
Azimuth angle {a} degrees
Stoomboot queue {q}
CORSIKA executable {c}
""".format(n=n, e=energy, p=particle, z=zenith, a=azimuth, q=queue,
c=corsika))
available_slots = check_queue(queue)
if available_slots <= 0:
n = 0
print 'Submitting no jobs because queue is full.'
return
elif available_slots < n:
n = available_slots
print 'Submitting {n} jobs because queue is almost full.'.format(n=n)
for _ in pbar(xrange(n)):
batch = CorsikaBatch(energy=energy, particle=particle, zenith=zenith,
azimuth=azimuth, queue=queue, corsika=corsika)
batch.run()
print 'Done.'
def update_sublevel_json(arg_type, type):
subdir = API.urls[type].split('/')[0]
try:
mkdir(path.join(LOCAL_BASE, subdir))
except OSError:
pass
url = API.urls[arg_type]
try:
numbers = [x['number'] for x in loads(API._retrieve_url(url))]
except:
print 'Failed to get %s data' % type
return
kwarg = API.urls[type].split('/')[1].strip('{}')
for number in pbar(numbers):
url = API.urls[type].format(**{kwarg: number,
'year': '', 'month': '', 'day': ''})
try:
get_and_store_json(url.strip('/'))
except:
print 'Failed to get %s data for %s %d' % (type, arg_type, number)
return
def serialiseddatafield_to_arrayfield(apps, schema_editor):
"""Forwards migrations"""
model = apps.get_model('histograms', 'DailyHistogram')
multi_model = apps.get_model('histograms', 'MultiDailyHistogram')
if model.objects.all().count():
print('')
for histogram in pbar(model.objects.all().iterator(), length=model.objects.all().count()):
if not histogram.type.has_multiple_datasets:
histogram.bins = histogram.old_bins
histogram.values = histogram.old_values
histogram.save()
else:
if histogram.type.slug == 'pulseheight':
new_bins = mv_to_adc(histogram.old_bins)
elif histogram.type.slug == 'pulseintegral':
new_bins = mvns_to_adcsample(histogram.old_bins)
else:
new_bins = histogram.old_bins
multi_model.objects.create(
source=histogram.source,
type=histogram.type,
bins=new_bins,
values=histogram.old_values)
histogram.delete()
def serialiseddatafield_to_arrayfield(apps, schema_editor):
"""Forwards migrations"""
model = apps.get_model('histograms', 'DailyHistogram')
multi_model = apps.get_model('histograms', 'MultiDailyHistogram')
if model.objects.all().count():
print('')
for histogram in pbar(model.objects.all().iterator(),
length=model.objects.all().count()):
if not histogram.type.has_multiple_datasets:
histogram.bins = histogram.old_bins
histogram.values = histogram.old_values
histogram.save()
else:
if histogram.type.slug == 'pulseheight':
new_bins = mv_to_adc(histogram.old_bins)
elif histogram.type.slug == 'pulseintegral':
new_bins = mvns_to_adcsample(histogram.old_bins)
else:
new_bins = histogram.old_bins
multi_model.objects.create(source=histogram.source,
type=histogram.type,
bins=new_bins,
values=histogram.old_values)
histogram.delete()
x = range(len(d1))
graph = Plot()
graph.plot(x, d1, markstyle='mark size=.5pt')
graph.plot(x, d2, markstyle='mark size=.5pt', linestyle='red')
graph.plot(x, d3, markstyle='mark size=.5pt', linestyle='green')
graph.plot(x, d4, markstyle='mark size=.5pt', linestyle='blue')
graph.set_ylabel('$\Delta t$ [ns]')
graph.set_xlabel('Date')
graph.set_xlimits(0, max(x))
graph.set_ylimits(-LIMITS, LIMITS)
graph.save_as_pdf('detector_offset_drift_%s_%d' % (type, station))
if __name__ == '__main__':
for station in pbar(STATIONS):
# Determine offsets for first day of each month
output = open('offsets_%d.tsv' % station, 'wb')
csvwriter = csv.writer(output, delimiter='\t')
offsets = []
timestamps = []
for y in range(2010, 2016):
for m in range(1, 13):
if y == 2015 and m >= 4:
continue
timestamps.append(datetime_to_gps(date(y, m, 1)))
path = os.path.join(DATA_PATH, str(y), str(m),
'%d_%d_1.h5' % (y, m))
with tables.open_file(path, 'r') as data:
offsets.append(determine_offset(data, station))
csvwriter.writerow([timestamps[-1]] + offsets[-1])
def download_events_data(data):
for station in pbar(STATIONS):
group = '/s%d' % station
if group not in data:
download_data(data, group, station, start=START, end=END,
progress=False)
def test_pbar_generator_wrong_length(self):
"""Raise exception for generator with wrong length"""
generator = (y for y in self.iterable)
pb = utils.pbar(generator, length=len(self.iterable) - 5, fd=self.output)
self.assertRaises(ValueError, list, pb)
self.n = len(self.files)
def __len__(self):
return self.n
def __iter__(self):
for fn in self.files:
with open(fn, 'rb') as f:
data = pickle.load(f)
yield data['station_id'], data['event_list']
if __name__ == '__main__':
datastore = Uploader(DATASTORE_VM, STATION_LIST)
folder_to_upload = sys.argv[1]
pickles = YieldPickles(folder_to_upload)
print('Uploading %d pickles from folder: %s' % (len(pickles),
folder_to_upload))
for sn, event_list in pbar(pickles):
assert sn in [98, 99]
try:
r = datastore.upload(sn, event_list)
except Exception as exc:
print('Connection failed: ', str(exc))
break
if r != '100':
print('Datastore responded with errorcoded: ', r, type(r))
break
Note: I normalized the event_ids in the events table. The event_ids were
1-based, and were modified to be 0-based. The following code was used:
events.modify_column(column=range(events.nrows), colname='event_id')
"""
import tables
from sapphire.utils import pbar
DATA_PATH = '/Users/arne/Datastore/kascade/kascade-20080912.h5'
if __name__ == "__main__":
ids = [1, 2, 3, 4]
with tables.open_file(DATA_PATH, 'r') as data:
event_ids = []
recs = data.root.reconstructions.iterrows()
events = data.root.hisparc.cluster_kascade.station_601.events.iterrows()
for rec in pbar(recs, length=data.root.reconstructions.nrows):
for event in events:
if all(rec['n%d' % id] == event['n%d' % id] for id in ids):
event_ids.append(event['event_id'])
break
with tables.open_file(DATA_PATH, 'a') as data:
data.create_array('/', 'c_index', event_ids)
events = data.root.hisparc.cluster_kascade.station_601.events
integrals = events.read_coordinates(event_ids, 'integrals')
data.create_array('/', 'reconstructions_integrals', integrals)
data.create_array('/', 'reconstructions_integrals_n', integrals / 5000.)
station_group = '/cluster_simulations/station_%d' % station.number
rec_events = ReconstructESDEvents(data,
station_group,
station,
overwrite=True,
progress=False)
rec_events.prepare_output()
rec_events.offsets = [d.offset for d in station.detectors]
rec_events.store_offsets()
rec_events.reconstruct_directions()
rec_events.store_reconstructions()
# Reconstruct coincidences
rec_coins = ReconstructESDCoincidences(data,
'/coincidences',
overwrite=True,
progress=False)
rec_coins.prepare_output()
rec_coins.offsets = {
station.number:
[d.offset + station.gps_offset for d in station.detectors]
for station in cluster.stations
}
rec_coins.reconstruct_directions()
rec_coins.store_reconstructions()
if __name__ == "__main__":
for path in pbar(glob.glob(PATHS)):
reconstruct_simulations(path)
def test_pbar_iterable(self):
pb = utils.pbar(self.iterable, fd=self.output)
self.assertIsInstance(pb, progressbar.ProgressBar)
self.assertEqual(list(pb), self.iterable)
def perform_simulations():
for id in pbar(range(10)):
script = SCRIPT.format(job_id=id)
submit_job(script, 'spa_sim_%d' % id, 'long')
