def test_event_descriptor_insertion():
# format some data keys for insertion
data_keys = {'some_value': {'source': 'PV:pv1',
'shape': [1, 2],
'dtype': 'array'},
'some_other_val': {'source': 'PV:pv2',
'shape': [],
'dtype': 'number'},
'data_key3': {'source': 'PV:pv1',
'shape': [],
'dtype': 'number',
'external': 'FS:foobar'}}
time = ttime.time()
# test insert
ev_desc_uid = mdsc.insert_descriptor(run_start_uid, data_keys, time,
str(uuid.uuid4()))
ev_desc_mds, = mdsc.find_descriptors(uid=ev_desc_uid)
# make sure the sanitized event descriptor has no uid
check_for_id(ev_desc_mds)
# make sure the event descriptor is pointing to the correct run start
referenced_run_start = ev_desc_mds['run_start']
assert referenced_run_start.uid == run_start_uid
assert ev_desc_mds['time'] == time
for k in data_keys:
for ik in data_keys[k]:
assert ev_desc_mds.data_keys[k][ik] == data_keys[k][ik]
def test_event_descriptor_insertion():
# format some data keys for insertion
data_keys = {'some_value': {'source': 'PV:pv1',
'shape': [1, 2],
'dtype': 'array'},
'some_other_val': {'source': 'PV:pv2',
'shape': [],
'dtype': 'number'},
'data_key3': {'source': 'PV:pv1',
'shape': [],
'dtype': 'number',
'external': 'FS:foobar'}}
time = ttime.time()
# test insert
ev_desc_uid = mdsc.insert_descriptor(run_start_uid, data_keys, time,
str(uuid.uuid4()))
ev_desc_mds, = mdsc.find_descriptors(uid=ev_desc_uid)
# make sure the sanitized event descriptor has no uid
check_for_id(ev_desc_mds)
# make sure the event descriptor is pointing to the correct run start
referenced_run_start = ev_desc_mds['run_start']
assert_equal(referenced_run_start.uid, run_start_uid)
assert_equal(ev_desc_mds['time'], time)
for k in data_keys:
for ik in data_keys[k]:
assert_equal(ev_desc_mds.data_keys[k][ik],
data_keys[k][ik])
def get_events(headers, fields=None, fill=True):
"""
Get Events from given run(s).
Parameters
----------
headers : Header or iterable of Headers
The headers to fetch the events for
fields : list, optional
whitelist of field names of interest; if None, all are returned
fill : bool, optional
Whether externally-stored data should be filled in. Defaults to True
Yields
------
event : Event
The event, optionally with non-scalar data filled in
"""
# A word about the 'fields' argument:
# Notice that we assume that the same field name cannot occur in
# more than one descriptor. We could relax this assumption, but
# we current enforce it in bluesky, so it is safe for now.
try:
headers.items()
except AttributeError:
pass
else:
headers = [headers]
if fields is None:
fields = []
fields = set(fields)
for header in headers:
descriptors = find_descriptors(header['start']['uid'])
for descriptor in descriptors:
all_fields = set(descriptor['data_keys'])
if fields:
discard_fields = all_fields - fields
else:
discard_fields = []
if discard_fields == all_fields:
continue
for event in get_events_generator(descriptor):
for field in discard_fields:
del event.data[field]
del event.timestamps[field]
if fill:
fill_event(event)
yield event
def test_find_events_smoke():
num = 50
rs, e_desc, data_keys = setup_syn()
all_data = syn_data(data_keys, num)
mdsc.bulk_insert_events(e_desc, all_data, validate=False)
mdsc.insert_run_stop(rs, ttime.time(), uid=str(uuid.uuid4()))
mdsc.clear_process_cache()
# make sure the uid works
next(mdsc.find_events(descriptor=e_desc))
mdsc.clear_process_cache()
descriptor, = mdsc.find_descriptors(uid=e_desc)
mdsc.clear_process_cache()
# make sure that searching by descriptor document works
next(mdsc.find_events(descriptor=descriptor))
def test_find_events_smoke():
num = 50
rs, e_desc, data_keys = setup_syn()
all_data = syn_data(data_keys, num)
mdsc.bulk_insert_events(e_desc, all_data, validate=False)
mdsc.insert_run_stop(rs, ttime.time(), uid=str(uuid.uuid4()))
mdsc.clear_process_cache()
# make sure the uid works
next(mdsc.find_events(descriptor=e_desc))
mdsc.clear_process_cache()
descriptor, = mdsc.find_descriptors(uid=e_desc)
mdsc.clear_process_cache()
# make sure that searching by descriptor document works
next(mdsc.find_events(descriptor=descriptor))
def get_table(headers, fields=None, fill=True, convert_times=True):
"""
Make a table (pandas.DataFrame) from given run(s).
Parameters
----------
headers : Header or iterable of Headers
The headers to fetch the events for
fields : list, optional
whitelist of field names of interest; if None, all are returned
fill : bool, optional
Whether externally-stored data should be filled in. Defaults to True
convert_times : bool, optional
Whether to convert times from float (seconds since 1970) to
numpy datetime64, using pandas. True by default.
Returns
-------
table : pandas.DataFrame
"""
# A word about the 'fields' argument:
# Notice that we assume that the same field name cannot occur in
# more than one descriptor. We could relax this assumption, but
# we current enforce it in bluesky, so it is safe for now.
try:
headers.items()
except AttributeError:
pass
else:
headers = [headers]
if fields is None:
fields = []
fields = set(fields)
dfs = []
for header in headers:
descriptors = find_descriptors(header['start']['uid'])
for descriptor in descriptors:
all_fields = set(descriptor['data_keys'])
if fields:
discard_fields = all_fields - fields
else:
discard_fields = []
if discard_fields == all_fields:
continue
is_external = _inspect_descriptor(descriptor)
payload = get_events_table(descriptor)
descriptor, data, seq_nums, times, uids, timestamps = payload
df = pd.DataFrame(index=seq_nums)
if convert_times:
times = pd.to_datetime(
pd.Series(times), unit='s', utc=True).dt.tz_localize(TZ)
df['time'] = times
for field, values in six.iteritems(data):
if field in discard_fields:
logger.debug('Discarding field %s', field)
continue
if is_external[field] and fill:
logger.debug('filling data for %s', field)
# TODO someday we will have bulk retrieve in FS
values = [fs.retrieve(value) for value in values]
df[field] = values
dfs.append(df)
if dfs:
return pd.concat(dfs)
else:
# edge case: no data
return pd.DataFrame()
def __call__(self, **kwargs):
"""Given search criteria, find Headers describing runs.
This function returns a list of dictionary-like objects encapsulating
the metadata for a run -- start time, instruments used, and so on.
In addition to the Parameters below, advanced users can specifiy
arbitrary queries that are passed through to mongodb.
Parameters
----------
start_time : time-like, optional
Include Headers for runs started after this time. Valid
"time-like" representations are:
- float timestamps (seconds since 1970), such as time.time()
- '2015'
- '2015-01'
- '2015-01-30'
- '2015-03-30 03:00:00'
- Python datetime objects, such as datetime.datetime.now()
stop_time: time-like, optional
Include Headers for runs started before this time. See
`start_time` above for examples.
beamline_id : str, optional
String identifier for a specific beamline
project : str, optional
Project name
owner : str, optional
The username of the logged-in user when the scan was performed
scan_id : int, optional
Integer scan identifier
uid : str, optional
Globally unique id string provided to metadatastore
data_key : str, optional
The alias (e.g., 'motor1') or PV identifier of data source
Returns
-------
data : list
Header objects
Examples
--------
>>> DataBroker(start_time='2015-03-05', stop_time='2015-03-10')
>>> DataBroker(data_key='motor1')
>>> DataBroker(data_key='motor1', start_time='2015-03-05')
"""
data_key = kwargs.pop('data_key', None)
run_start = find_run_starts(**kwargs)
if data_key is not None:
node_name = 'data_keys.{0}'.format(data_key)
query = {node_name: {'$exists': True}}
descriptors = []
for rs in run_start:
descriptor = find_descriptors(run_start=rs, **query)
for d in descriptor:
descriptors.append(d)
# query = {node_name: {'$exists': True},
# 'run_start_id': {'$in': [ObjectId(rs.id) for rs in run_start]}}
# descriptors = find_descriptors(**query)
result = []
known_uids = deque()
for descriptor in descriptors:
if descriptor['run_start']['uid'] not in known_uids:
rs = descriptor['run_start']
known_uids.append(rs['uid'])
result.append(rs)
run_start = result
result = []
for rs in run_start:
result.append(Header.from_run_start(rs))
return result
