def run_steps(self, N, progress_bar=True):
if self.closed:
raise SimulatorClosed("Simulator cannot run because it is closed.")
if self.n_steps + N >= 2**24:
# since n_steps is float32, point at which `n_steps == n_steps + 1`
raise ValueError("Cannot handle more than 2**24 steps")
if self._cl_probe_plan is not None:
# -- precondition: the probe buffers have been drained
bufpositions = self._cl_probe_plan.cl_bufpositions.get()
assert np.all(bufpositions == 0)
if progress_bar is None:
progress_bar = self.progress_bar
try:
progress = ProgressTracker(N, progress_bar, "Simulating")
except TypeError:
progress = ProgressTracker(N, progress_bar)
with progress:
# -- we will go through N steps of the simulator
# in groups of up to B at a time, draining
# the probe buffers after each group of B
while N:
B = min(N, self._max_steps_between_probes)
self._plans.call_n_times(B)
self._probe()
N -= B
progress.step(n=B)
if self.profiling > 1:
self.print_profiling()
def test_progress_tracker():
update_interval = 0.001
sleep_interval = 20 * update_interval
stages = 4
steps = 3
total_progress = Progress(name_during="total_prog", max_steps=stages)
progress_bar = ProgressBarMock()
tracker = ProgressTracker(progress_bar,
total_progress,
update_interval=update_interval)
assert not progress_bar.closed
with tracker:
for i in range(stages):
sub_progress = tracker.next_stage(name_during="stage%d" % i,
max_steps=steps)
with sub_progress:
for j in range(steps):
sub_progress.step()
time.sleep(sleep_interval)
assert progress_bar.updates[
-1].name_during == "stage%d" % i
assert progress_bar.updates[-1].n_steps == j + 1
time.sleep(sleep_interval)
assert progress_bar.updates[-1].name_during == "total_prog"
assert progress_bar.updates[-1].n_steps == i
assert progress_bar.closed
def __init__(self,
network,
dt=0.001,
seed=None,
model=None,
progress_bar=True,
optimize=True):
self.closed = True # Start closed in case constructor raises exception
self.progress_bar = progress_bar
self.optimize = optimize
if model is None:
self.model = Model(
dt=float(dt),
label="%s, dt=%f" % (network, dt),
decoder_cache=get_default_decoder_cache(),
)
else:
self.model = model
pt = ProgressTracker(progress_bar, Progress("Building", "Build"))
with pt:
if network is not None:
# Build the network into the model
self.model.build(network,
progress=pt.next_stage("Building", "Build"))
# Order the steps (they are made in `Simulator.reset`)
self.dg = operator_dependency_graph(self.model.operators)
if optimize:
with pt.next_stage("Building (running optimizer)",
"Optimization"):
opmerge_optimize(self.model, self.dg)
self._step_order = [
op for op in toposort(self.dg) if hasattr(op, "make_step")
]
# -- map from Signal.base -> ndarray
self.signals = SignalDict()
for op in self.model.operators:
op.init_signals(self.signals)
# Add built states to the raw simulation data dictionary
self._sim_data = self.model.params
# Provide a nicer interface to simulation data
self.data = SimulationData(self._sim_data)
if seed is None:
if network is not None and network.seed is not None:
seed = network.seed + 1
else:
seed = np.random.randint(npext.maxint)
self.closed = False
self.reset(seed=seed)
def run_steps(self, steps, progress_bar=None):
"""Simulate for the given number of ``dt`` steps.
Parameters
----------
steps : int
Number of steps to run the simulation for.
progress_bar : bool or `.ProgressBar` or `.ProgressUpdater`, optional \
(Default: True)
Progress bar for displaying the progress of the simulation run.
If True, the default progress bar will be used.
If False, the progress bar will be disabled.
For more control over the progress bar, pass in a `.ProgressBar`
or `.ProgressUpdater` instance.
"""
if progress_bar is None:
progress_bar = self.progress_bar
with ProgressTracker(progress_bar,
Progress("Simulating", "Simulation",
steps)) as pt:
for i in range(steps):
self.step()
pt.total_progress.step()
def publish_loop(self, steps, progress_bar=True):
manager = yield From(pygazebo.connect())
def callback_pose(data):
self.sensor_comp.callback(data)
publisher = yield From(
manager.advertise('/gazebo/default/turtlebot/joint_cmd',
'gazebo.msgs.JointCmd'))
subscriber_pose = manager.subscribe('/gazebo/default/pose/info',
'gazebo.msgs.PosesStamped',
callback_pose)
yield From(subscriber_pose.wait_for_connection())
msg_left_wheel = pygazebo.msg.joint_cmd_pb2.JointCmd()
msg_left_wheel.name = 'turtlebot::create::left_wheel'
msg_right_wheel = pygazebo.msg.joint_cmd_pb2.JointCmd()
msg_right_wheel.name = 'turtlebot::create::right_wheel'
print "wait for connection.."
yield From(trollius.sleep(1.0))
with ProgressTracker(steps, progress_bar) as progress:
for i in range(steps):
self.step()
force = self.action_comp.callback()
max_force_strength = 0.7
force = np.clip(force, -max_force_strength, max_force_strength)
msg_right_wheel.force, msg_left_wheel.force = force
From(publisher.publish(msg_left_wheel))
From(publisher.publish(msg_right_wheel))
yield From(trollius.sleep(0.01))
def test_at_most_n_updates_are_performed(self):
progress_bar = ProgressBarMock()
updater = UpdateN(progress_bar, max_updates=3)
with ProgressTracker(100, updater) as p:
for _ in range(100):
p.step()
assert progress_bar.n_update_calls > 0
assert progress_bar.n_update_calls <= 3
def run_steps(self, N, progress_bar=True):
if self.closed:
raise SimulatorClosed("Simulator cannot run because it is closed.")
if self.n_steps + N >= 2**24:
# since n_steps is float32, point at which `n_steps == n_steps + 1`
raise ValueError("Cannot handle more than 2**24 steps")
if self._cl_probe_plan is not None:
# -- precondition: the probe buffers have been drained
bufpositions = self._cl_probe_plan.cl_bufpositions.get()
assert np.all(bufpositions == 0)
if progress_bar is None:
progress_bar = self.progress_bar
try:
progress = ProgressTracker(N, progress_bar, "Simulating")
except TypeError:
progress = ProgressTracker(N, progress_bar)
with progress:
# -- we will go through N steps of the simulator
# in groups of up to B at a time, draining
# the probe buffers after each group of B
while N:
B = min(N, self._max_steps_between_probes)
self._plans.call_n_times(B)
self._probe()
N -= B
progress.step(n=B)
if self.profiling > 1:
self.print_profiling()
def test_updates_every_n_steps(self):
progress_bar = ProgressBarMock()
updater = UpdateEveryN(progress_bar, every_n=5)
with ProgressTracker(100, updater) as p:
progress_bar.n_update_calls = 0
for _ in range(5):
p.step()
assert progress_bar.n_update_calls == 1
p.step(2)
assert progress_bar.n_update_calls == 1
p.step(3)
assert progress_bar.n_update_calls == 2
def test_updates_after_interval_has_passed(self, monkeypatch):
progress_bar = ProgressBarMock()
updater = UpdateEveryT(progress_bar, every_t=2.)
t = 1.
monkeypatch.setattr(time, 'time', lambda: t)
with ProgressTracker(100, updater) as p:
p.step() # Update is allowed to happen on first step.
progress_bar.n_update_calls = 0
p.step()
assert progress_bar.n_update_calls == 0
t = 2.
p.step()
assert progress_bar.n_update_calls == 0
t = 4.
p.step()
assert progress_bar.n_update_calls == 1
def run_steps(self, N, progress_bar=True):
has_probes = self._cl_probe_plan is not None
if has_probes:
# -- precondition: the probe buffers have been drained
bufpositions = self._cl_probe_plan.cl_bufpositions.get()
assert np.all(bufpositions == 0)
with ProgressTracker(N, progress_bar) as progress:
# -- we will go through N steps of the simulator
# in groups of up to B at a time, draining
# the probe buffers after each group of B
while N:
B = min(N, self._max_steps_between_probes)
self._plans.call_n_times(B)
if has_probes:
self.drain_probe_buffers()
N -= B
self.n_steps += B
progress.step(n=B)
if self.profiling > 1:
self.print_profiling()
def run_steps(self, steps, progress_bar=True):
"""Simulate for the given number of `dt` steps.
Parameters
----------
steps : int
Number of steps to run the simulation for.
progress_bar : bool or ``ProgressBar`` or ``ProgressUpdater``, optional
Progress bar for displaying the progress.
By default, ``progress_bar=True``, which uses the default progress
bar (text in most situations, or an HTML version in recent IPython
notebooks).
To disable the progress bar, use ``progress_bar=False``.
For more control over the progress bar, pass in a
:class:`nengo.utils.progress.ProgressBar`,
or :class:`nengo.utils.progress.ProgressUpdater` instance.
"""
with ProgressTracker(steps, progress_bar) as progress:
for i in range(steps):
self.step()
progress.step()
def build_network(model, network, progress_bar=False):
"""Builds a `.Network` object into a model.
The network builder does this by mapping each high-level object to its
associated signals and operators one-by-one, in the following order:
1. Ensembles, nodes, neurons
2. Subnetworks (recursively)
3. Connections, learning rules
4. Probes
Before calling any of the individual objects' build functions, random
number seeds are assigned to objects that did not have a seed explicitly
set by the user. Whether the seed was assigned manually or automatically
is tracked, and the decoder cache is only used when the seed is assigned
manually.
Parameters
----------
model : Model
The model to build into.
network : Network
The network to build.
progress_bar : bool or `.ProgressBar` or `.ProgressUpdater`, optional \
(Default: False)
Progress bar for displaying build progress.
If True, the default progress bar will be used.
If False, the progress bar will be disabled.
For more control over the progress bar, pass in a `.ProgressBar`
or `.ProgressUpdater` instance.
Note that this will only affect top-level networks. Subnetworks
cannot have progress bars displayed.
Notes
-----
Sets ``model.params[network]`` to ``None``.
"""
def get_seed(obj, rng):
# Generate a seed no matter what, so that setting a seed or not on
# one object doesn't affect the seeds of other objects.
seed = rng.randint(npext.maxint)
return (seed if not hasattr(obj, 'seed') or obj.seed is None
else obj.seed)
if model.toplevel is None:
model.toplevel = network
model.seeds[network] = get_seed(network, np.random)
model.seeded[network] = getattr(network, 'seed', None) is not None
else:
progress_bar = False
max_steps = max(1, len(network.all_objects))
progress = ProgressTracker(max_steps, progress_bar, task="Building")
# Set config
old_config = model.config
model.config = network.config
# assign seeds to children
rng = np.random.RandomState(model.seeds[network])
sorted_types = sorted(network.objects, key=lambda t: t.__name__)
for obj_type in sorted_types:
for obj in network.objects[obj_type]:
model.seeded[obj] = (model.seeded[network] or
getattr(obj, 'seed', None) is not None)
model.seeds[obj] = get_seed(obj, rng)
# If this is the toplevel network, enter the decoder cache
context = (model.decoder_cache if model.toplevel is network
else nullcontext())
with context, progress:
model.build_callback = lambda obj: progress.step()
logger.debug("Network step 1: Building ensembles and nodes")
for obj in network.ensembles + network.nodes:
model.build(obj)
logger.debug("Network step 2: Building subnetworks")
for subnetwork in network.networks:
model.build(subnetwork)
logger.debug("Network step 3: Building connections")
for conn in network.connections:
# NB: we do these in the order in which they're defined, and build
# the learning rule in the connection builder. Because learning
# rules are attached to connections, the connection that contains
# the learning rule (and the learning rule) are always built
# *before* a connection that attaches to that learning rule.
# Therefore, we don't have to worry about connection ordering here.
# TODO: Except perhaps if the connection being learned
# is in a subnetwork?
model.build(conn)
logger.debug("Network step 4: Building probes")
for probe in network.probes:
model.build(probe)
if context is model.decoder_cache:
model.decoder_cache.shrink()
model.build_callback = None
# Unset config
model.config = old_config
model.params[network] = None
def build_network(model, network, progress_bar=False):
"""Builds a `.Network` object into a model.
The network builder does this by mapping each high-level object to its
associated signals and operators one-by-one, in the following order:
1. Ensembles, nodes, neurons
2. Subnetworks (recursively)
3. Connections, learning rules
4. Probes
Before calling any of the individual objects' build functions, random
number seeds are assigned to objects that did not have a seed explicitly
set by the user. Whether the seed was assigned manually or automatically
is tracked, and the decoder cache is only used when the seed is assigned
manually.
Parameters
----------
model : Model
The model to build into.
network : Network
The network to build.
progress_bar : bool or `.ProgressBar` or `.ProgressUpdater`, optional \
(Default: False)
Progress bar for displaying build progress.
If True, the default progress bar will be used.
If False, the progress bar will be disabled.
For more control over the progress bar, pass in a `.ProgressBar`
or `.ProgressUpdater` instance.
Note that this will only affect top-level networks. Subnetworks
cannot have progress bars displayed.
Notes
-----
Sets ``model.params[network]`` to ``None``.
"""
def get_seed(obj, rng):
# Generate a seed no matter what, so that setting a seed or not on
# one object doesn't affect the seeds of other objects.
seed = rng.randint(npext.maxint)
return (seed
if not hasattr(obj, 'seed') or obj.seed is None else obj.seed)
if model.toplevel is None:
model.toplevel = network
model.seeds[network] = get_seed(network, np.random)
model.seeded[network] = getattr(network, 'seed', None) is not None
else:
progress_bar = False
max_steps = len(network.all_objects) + 1 # +1 for top level network itself
progress = ProgressTracker(max_steps, progress_bar, task="Building")
# Set config
old_config = model.config
model.config = network.config
# assign seeds to children
rng = np.random.RandomState(model.seeds[network])
# Put probes last so that they don't influence other seeds
sorted_types = (Connection, Ensemble, Network, Node, Probe)
assert all(tp in sorted_types for tp in network.objects)
for obj_type in sorted_types:
for obj in network.objects[obj_type]:
model.seeded[obj] = (model.seeded[network]
or getattr(obj, 'seed', None) is not None)
model.seeds[obj] = get_seed(obj, rng)
# If this is the toplevel network, enter the decoder cache
context = (model.decoder_cache
if model.toplevel is network else nullcontext())
with context, progress:
def build_callback(obj):
if isinstance(obj, tuple(network.objects)):
progress.step()
model.build_callback = build_callback
logger.debug("Network step 1: Building ensembles and nodes")
for obj in network.ensembles + network.nodes:
model.build(obj)
logger.debug("Network step 2: Building subnetworks")
for subnetwork in network.networks:
model.build(subnetwork)
logger.debug("Network step 3: Building connections")
for conn in network.connections:
# NB: we do these in the order in which they're defined, and build
# the learning rule in the connection builder. Because learning
# rules are attached to connections, the connection that contains
# the learning rule (and the learning rule) are always built
# *before* a connection that attaches to that learning rule.
# Therefore, we don't have to worry about connection ordering here.
# TODO: Except perhaps if the connection being learned
# is in a subnetwork?
model.build(conn)
logger.debug("Network step 4: Building probes")
for probe in network.probes:
model.build(probe)
if context is model.decoder_cache:
model.decoder_cache.shrink()
progress.step()
model.build_callback = None
# Unset config
model.config = old_config
model.params[network] = None
