def to_operation(self, inputs, conditions={}):
outputs = self.output_vars()
def fit(arr):
arr = np.array(arr)
slope, intercept, r_value, p_value, stderr = stats.linregress(
arr[:, 0], arr[:, 1])
return arr[:, 0], arr[:, 1], slope * arr[:, 0] + intercept, r_value
picked_outputs = [self.name() + "_accumulated"]
nodes = [
gn.PickN(name=self.name() + "_picked",
condition_needs=conditions,
inputs=inputs,
outputs=picked_outputs,
N=self.values['N'],
parent=self.name()),
gn.Map(name=self.name() + "_operation",
inputs=picked_outputs,
outputs=outputs,
func=fit,
parent=self.name())
]
return nodes
def _collect_global_inputs(self):
"""
Insert Pick1 for nodes which run global collector but depend on inputs which are only available on worker.
"""
inputs = [n for n, d in self.graph.in_degree() if d == 0]
global_collector_nodes = list(
filter(
lambda node: getattr(node, 'color', '') == 'globalCollector',
self.graph.nodes))
for node in global_collector_nodes:
new_inputs = []
update_inputs = False
if node in self.global_operations:
continue
for i in node.inputs:
if i in inputs:
pickone = gn.PickN(name=i + "_pick1",
inputs=[i],
outputs=["one_" + i],
parent=node.parent)
self.global_operations.add(pickone)
self.add(pickone)
update_inputs = True
new_inputs.extend(pickone.outputs)
else:
new_inputs.append(i)
self.graph.remove_node(node)
if update_inputs:
node.inputs = new_inputs
self.add(node)
def to_operation(self, inputs, conditions={}):
outputs = self.output_vars()
node = gn.PickN(name=self.name() + "_operation",
inputs=inputs,
outputs=outputs,
condition_needs=conditions,
N=self.values['N'],
parent=self.name())
return node
def to_operation(self, inputs, conditions={}):
outputs = self.output_vars()
axis = self.values['axis']
accumulated_outputs = [self.name() + '_accumulated_events']
if self.values['infinite']:
def reduction(res, *rest):
if type(rest[0]) is list:
res[0] = np.sum([res[0], rest[0][0]], axis=axis)
res[1] += rest[0][1]
else:
res[0] = np.sum([res[0], rest[0]], axis=axis)
res[1] = res[1] + 1
return res
def avg(*args, **kwargs):
return args[0][0] / args[0][1]
nodes = [
gn.Accumulator(name=self.name() + "_accumulated",
inputs=inputs,
outputs=accumulated_outputs,
condition_needs=conditions,
res_factory=lambda: [0, 0],
reduction=reduction,
parent=self.name()),
gn.Map(name=self.name() + "_map",
inputs=accumulated_outputs,
outputs=outputs,
func=avg,
parent=self.name())
]
else:
def func(arrs):
return np.sum(arrs, axis=axis) / len(arrs)
nodes = [
gn.PickN(name=self.name() + "_accumulated",
inputs=inputs,
outputs=accumulated_outputs,
condition_needs=conditions,
N=self.values['N'],
parent=self.name()),
gn.Map(name=self.name() + "_map",
inputs=accumulated_outputs,
outputs=outputs,
func=func,
parent=self.name())
]
return nodes
def to_operation(self, inputs, outputs, **kwargs):
accumulated_outputs = [self.name() + '_accumulated_events']
if self.values['infinite']:
def reduction(res, *rest):
if type(rest[0]) is list:
res[0] += rest[0][0]
res[1] += rest[0][1]
else:
res[0] = res[0] + rest[0]
res[1] = res[1] + 1
return res
def avg(*args, **kwargs):
return args[0][0] / args[0][1]
nodes = [
gn.Accumulator(name=self.name() + "_accumulated",
inputs=inputs,
outputs=accumulated_outputs,
res_factory=lambda: [0, 0],
reduction=reduction,
**kwargs),
gn.Map(name=self.name() + "_map",
inputs=accumulated_outputs,
outputs=outputs,
func=avg,
**kwargs)
]
else:
def func(arrs):
return np.average(arrs)
nodes = [
gn.PickN(name=self.name() + "_accumulated",
inputs=inputs,
outputs=accumulated_outputs,
N=self.values['N'],
**kwargs),
gn.Map(name=self.name() + "_map",
inputs=accumulated_outputs,
outputs=outputs,
func=func,
**kwargs)
]
return nodes
def to_operation(self, inputs, outputs, **kwargs):
outputs = [self.name()+"_unused"]
picked_outputs = [self.name()+"_pickedoutput"]
if self.values['global']:
nodes = [gn.PickN(name=self.name()+"_picked",
inputs=inputs, outputs=picked_outputs, N=1, **kwargs),
gn.Map(name=self.name()+"_operation",
inputs=picked_outputs, outputs=outputs,
func=CaputProc(self.values['pvname']), **kwargs)]
else:
nodes = [gn.Map(name=self.name()+"_operation",
inputs=inputs, outputs=outputs,
func=CaputProc(self.values['pvname']), **kwargs)]
return nodes
def to_operation(self, inputs, outputs, **kwargs):
map_outputs = [self.name() + "_map"]
threshold = self.values['Threshold']
count = self.values['Count']
nodes = [
gn.Map(name=self.name() + "_operation",
inputs=inputs,
outputs=map_outputs,
**kwargs,
func=lambda arr: len(arr[arr > threshold]) > count),
gn.PickN(name=self.name() + "_pickN",
inputs=map_outputs,
outputs=outputs,
**kwargs)
]
return nodes
def to_operation(self, inputs, conditions={}):
outputs = self.output_vars()
pickn_outputs = [self.name() + "_picked"]
display_outputs = [
self.name() + "_displayX",
self.name() + "_displayY"
]
def display_func(arr):
x, y = zip(*arr)
return np.array(x), np.array(y)
origin = self.values['origin']
extent = self.values['extent']
def func(arr):
arr = np.array(arr)
roi = arr[(origin < arr[:, 0]) & (arr[:, 0] < extent)]
if roi.size > 0:
return roi[:, 0], roi[:, 1]
else:
return np.array([]), np.array([])
nodes = [
gn.PickN(name=self.name() + "_pickN",
condition_needs=conditions,
inputs=inputs,
outputs=pickn_outputs,
parent=self.name(),
N=self.values['Num Points']),
gn.Map(name=self.name() + "_operation",
inputs=pickn_outputs,
outputs=outputs,
func=func,
parent=self.name()),
gn.Map(name=self.name() + "_display",
inputs=pickn_outputs,
outputs=display_outputs,
parent=self.name(),
func=display_func)
]
return nodes
def to_operation(self, inputs, conditions={}):
map_outputs = [self.name() + "_map"]
outputs = [self.name()]
threshold = self.values['Threshold']
count = self.values['Count']
nodes = [
gn.Map(name=self.name() + "_operation",
condition_needs=conditions,
inputs=inputs,
outputs=map_outputs,
func=lambda arr: len(arr[arr > threshold]) > count,
parent=self.name()),
gn.PickN(name=self.name() + "_pickN",
inputs=map_outputs,
outputs=outputs,
parent=self.name())
]
return nodes
def to_operation(self, inputs, outputs, **kwargs):
def func(arr):
mean = np.mean(arr)
rms = np.sqrt(np.mean(np.square(arr)))
return mean, rms
accumulated_outputs = [self.name() + '_accumulated_events']
nodes = [
gn.PickN(name=self.name() + '_picked',
N=self.values['N'],
inputs=inputs,
outputs=accumulated_outputs,
**kwargs),
gn.Map(name=self.name() + '_operation',
inputs=accumulated_outputs,
outputs=outputs,
func=func,
**kwargs)
]
return nodes
def to_operation(self, inputs, outputs, **kwargs):
def fit(arr):
arr = np.array(arr)
slope, intercept, r_value, p_value, stderr = stats.linregress(
arr[:, 0], arr[:, 1])
return arr[:, 0], arr[:, 1], slope * arr[:, 0] + intercept, r_value
picked_outputs = [self.name() + "_accumulated"]
nodes = [
gn.PickN(name=self.name() + "_picked",
inputs=inputs,
outputs=picked_outputs,
N=self.values['N'],
**kwargs),
gn.Map(name=self.name() + "_operation",
inputs=picked_outputs,
outputs=outputs,
func=fit,
**kwargs)
]
return nodes
def test_pickn(qtbot):
node = PickN('pickn')
widget = node.ctrlWidget()
# widget.show()
qtbot.addWidget(widget)
assert node.values['N'] == 2
qtbot.keyPress(node.ctrls['N'], QtCore.Qt.Key_Up)
assert node.values['N'] == 3
inputs = {"In": node.name()}
outputs = ['pickn.Out']
op = node.to_operation(inputs=inputs, outputs=outputs)
pop = gn.PickN(name="pickn_operation",
inputs=list(inputs.values()),
outputs=[node.name() + '.Out'],
N=3)
assert op.name == pop.name
assert op.inputs == pop.inputs
assert op.outputs == pop.outputs
assert op.N == pop.N
def to_operation(self, **kwargs):
return gn.PickN(name=self.name() + "_operation",
N=self.values['N'],
**kwargs)
def to_operation(self, **kwargs):
return gn.PickN(name=self.name() + "_operation", N=1, **kwargs)