def func():
m = motor.read()[motor_field]['value']
v = Imax * np.exp(-(m - center)**2 / (2 * sigma**2))
arr = self.gauss(dims,
img_sigma) * v + np.random.random(dims) * .01
fs_uid = save_ndarray(arr, self.output_dir)
return fs_uid
def trigger(self, *, block_group=True):
self.ready = False
m = self._motor._data[self._motor_field]['value']
v = self.Imax * np.exp(-(m - self.center)**2 / (2 * self.sigma**2))
arr = self.gauss(self.dims, self.img_sigma) * v + np.random.random(
self.dims) * .01
fs_uid = save_ndarray(arr, self.output_dir)
self._data = {self._name: {'value': fs_uid, 'timestamp': ttime.time()}}
ttime.sleep(0.05) # simulate exposure time
self.ready = True
return self
def run(run_start=None, sleep=0):
# Make the data
rs = np.random.RandomState(5)
# set up the data keys entry
data_keys1 = {'linear_motor': dict(source='PV:ES:sam_x', dtype='number'),
'img': dict(source='CCD', shape=(5, 5), dtype='array',
external='FILESTORE:'),
'total_img_sum': dict(source='CCD:sum', dtype='number'),
'img_x_max': dict(source='CCD:xmax', dtype='number'),
'img_y_max': dict(source='CCD:ymax', dtype='number'),
'img_sum_x': dict(source='CCD:xsum', dtype='array',
shape=(5,), external='FILESTORE:'),
'img_sum_y': dict(source='CCD:ysum', dtype='array',
shape=(5,), external='FILESTORE:')
}
data_keys2 = {'Tsam': dict(source='PV:ES:Tsam', dtype='number')}
# save the first event descriptor
e_desc1 = insert_event_descriptor(
run_start=run_start, data_keys=data_keys1, time=0.,
uid=str(uuid.uuid4()))
e_desc2 = insert_event_descriptor(
run_start=run_start, data_keys=data_keys2, time=0.,
uid=str(uuid.uuid4()))
# number of motor positions to fake
num1 = 20
# number of temperatures to record per motor position
num2 = 10
events = []
for idx1, i in enumerate(range(num1)):
img = next(frame_generator)
img_sum = float(img.sum())
img_sum_x = img.sum(axis=0)
img_sum_y = img.sum(axis=1)
img_x_max = float(img_sum_x.argmax())
img_y_max = float(img_sum_y.argmax())
fsid_img = save_ndarray(img)
fsid_x = save_ndarray(img_sum_x)
fsid_y = save_ndarray(img_sum_y)
# Put in actual ndarray data, as broker would do.
data1 = {'linear_motor': (i, noisy(i)),
'total_img_sum': (img_sum, noisy(i)),
'img': (fsid_img, noisy(i)),
'img_sum_x': (fsid_x, noisy(i)),
'img_sum_y': (fsid_y, noisy(i)),
'img_x_max': (img_x_max, noisy(i)),
'img_y_max': (img_y_max, noisy(i))
}
event = insert_event(event_descriptor=e_desc1, seq_num=idx1,
time=noisy(i), data=data1, uid=str(uuid.uuid4()))
fill_event(event)
events.append(event)
for idx2, i2 in enumerate(range(num2)):
time = noisy(i/num2)
data2 = {'Tsam': (idx1 + np.random.randn()/100, time)}
event = insert_event(event_descriptor=e_desc2, seq_num=idx2+idx1,
time=time, data=data2, uid=str(uuid.uuid4()))
events.append(event)
ttime.sleep(sleep)
return events
def func():
m = motor.read()[motor_field]['value']
v = Imax * np.exp(-(m - center)**2 / (2 * sigma**2))
arr = self.gauss(dims, img_sigma) * v + np.random.random(dims) * .01
fs_uid = save_ndarray(arr, self.output_dir)
return fs_uid
def run(run_start_uid=None, sleep=0):
frame_generator = frame_generators.brownian(img_size, step_scale=.5,
I_fluc_function=I_func_gaus,
step_fluc_function=scale_fluc)
# seed data to make deterministic
np.random.RandomState(5)
# set up the data keys entry
data_keys1 = {'linear_motor': dict(source='PV:ES:sam_x', dtype='number'),
'img': dict(source='CCD', shape=(5, 5), dtype='array',
external='FILESTORE:'),
'total_img_sum': dict(source='CCD:sum', dtype='number'),
'img_x_max': dict(source='CCD:xmax', dtype='number'),
'img_y_max': dict(source='CCD:ymax', dtype='number'),
'img_sum_x': dict(source='CCD:xsum', dtype='array',
shape=(5,), external='FILESTORE:'),
'img_sum_y': dict(source='CCD:ysum', dtype='array',
shape=(5,), external='FILESTORE:')
}
data_keys2 = {'Tsam': dict(source='PV:ES:Tsam', dtype='number')}
# save the first event descriptor
descriptor1_uid = insert_descriptor(
run_start=run_start_uid, data_keys=data_keys1, time=0.,
uid=str(uuid.uuid4()))
descriptor2_uid = insert_descriptor(
run_start=run_start_uid, data_keys=data_keys2, time=0.,
uid=str(uuid.uuid4()))
events = []
for idx1, i in enumerate(range(num1)):
img = next(frame_generator)
img_sum = float(img.sum())
img_sum_x = img.sum(axis=0)
img_sum_y = img.sum(axis=1)
img_x_max = float(img_sum_x.argmax())
img_y_max = float(img_sum_y.argmax())
fsid_img = save_ndarray(img)
fsid_x = save_ndarray(img_sum_x)
fsid_y = save_ndarray(img_sum_y)
# Put in actual ndarray data, as broker would do.
data1 = {'linear_motor': i,
'total_img_sum': img_sum,
'img': fsid_img,
'img_sum_x': fsid_x,
'img_sum_y': fsid_y,
'img_x_max': img_x_max,
'img_y_max': img_y_max
}
timestamps1 = {k: noisy(i) for k in data1}
event_uid = insert_event(descriptor=descriptor1_uid, seq_num=idx1,
time=noisy(i), data=data1,
timestamps=timestamps1,
uid=str(uuid.uuid4()))
event, = find_events(uid=event_uid)
events.append(event)
for idx2, i2 in enumerate(range(num2)):
time = noisy(i/num2)
data2 = {'Tsam': idx1 + np.random.randn()}
timestamps2 = {'Tsam': time}
event_uid = insert_event(descriptor=descriptor2_uid,
seq_num=idx2+idx1, time=time, data=data2,
uid=str(uuid.uuid4()),
timestamps=timestamps2)
event, = find_events(uid=event_uid)
events.append(event)
ttime.sleep(sleep)
return events
def _npsave_helper(dd, base_path):
eid = fs_write.save_ndarray(dd, base_path)
with fsa.handler_context({'npy': fs_read.NpyHandler}):
ret = fsa.retrieve(eid)
assert_array_equal(dd, ret)