How can I stop outputing "spikerates"
Xingyu DENG
Hi, I am Prof. Steve Furber’s student, and I am doing my project.
I would like to ask a question about how can I stop outputting “spikerate” when running simulation on spiNNaker on snn toolbox.
Here is my configuration of snn toolbox simulation (except red highlighted).
“
configparser = import_configparser()
config = configparser.ConfigParser()
config['paths'] = {
'path_wd': path_wd, # Path to model.
'dataset_path': path_wd, # Path to dataset.
'filename_ann': model_name # Name of input model.
}
config['tools'] = {
'evaluate_ann': True, # Test ANN on dataset before conversion.
'normalize': True, # Normalize weights for full dynamic range.
}
config['simulation'] = {
'simulator': 'spiNNaker',
'duration' : '200',
'dt' : '1',
'batch_size' : '1',
'num_to_test' : '10000',
}
config['input'] = {
'poisson_input': True # Images are encodes as spike trains.
}
config['output'] = {
'plot_vars': { # Various plots (slows down simulation).
'spikerates',
'error_t'}
}
“
When I run the simulation, there is error called “ConfigurationException: This population has not been set to record spikes” (I attached full error output of it at the end)
I changed parameter 'simulator' of simulation configuration from ‘spiNNaker’ to ‘INI’, then it works.
I added parameter 'spikerates' of output configuration above, then it works.
Hence, I believe that the error only happens when I use spiNNaker as simulator without outputting ‘spikerates’.
How can I solve it?
The spiNNaker on toolbox only support sequential test, which means it outputs a result for each test sample.
It unpractical if I have to output figures of “spikerates” for each test sample.
Thank you very much in advance!
Xingyu :D
---------------------------------------------------------------------------
ConfigurationException Traceback (most recent call last)
<ipython-input-2-14d6fb19a662> in <module>
2 ###################
3
----> 4 main(config_filepath)
~/sPyNNaker/lib/python3.8/site-packages/snntoolbox/bin/run.py in main(filepath)
29 if filepath is not None:
30 config = update_setup(filepath)
---> 31 run_pipeline(config)
32 return
33
~/sPyNNaker/lib/python3.8/site-packages/snntoolbox/bin/utils.py in run_pipeline(config, queue)
143
144 # Simulate network
--> 145 results = run(spiking_model, **testset)
146
147 # Clean up
~/sPyNNaker/lib/python3.8/site-packages/snntoolbox/bin/utils.py in wrapper(snn, **testset)
218 "{} = {:.2f}\n".format(param_name, p))
219
--> 220 results.append(run_single(snn, **testset))
221
222 # Plot and return results of parameter sweep.
~/sPyNNaker/lib/python3.8/site-packages/snntoolbox/bin/utils.py in run(snn, **test_set)
140 @run_parameter_sweep(config, queue)
141 def run(snn, **test_set):
--> 142 return snn.run(**test_set)
143
144 # Simulate network
~/sPyNNaker/lib/python3.8/site-packages/snntoolbox/simulation/utils.py in run(self, x_test, y_test, dataflow, **kwargs)
604 # of the simulation.
605 print("\nStarting new simulation...\n")
--> 606 output_b_l_t = self.simulate(**data_batch_kwargs)
607
608 # Halt if model is to be serialised only.
~/sPyNNaker/lib/python3.8/site-packages/snntoolbox/simulation/target_simulators/spiNNaker_target_sim.py in simulate(self, **kwargs)
381 self.sim.run(self._duration)
382 print("\nCollecting results...")
--> 383 output_b_l_t = self.get_recorded_vars(self.layers)
384
385 return output_b_l_t
~/sPyNNaker/lib/python3.8/site-packages/snntoolbox/simulation/utils.py in get_recorded_vars(self, layers)
960 """
961
--> 962 self.set_spiketrain_stats_input()
963
964 self.reset_container_counters()
~/sPyNNaker/lib/python3.8/site-packages/snntoolbox/simulation/target_simulators/pyNN_target_sim.py in set_spiketrain_stats_input(self)
514
515 def set_spiketrain_stats_input(self):
--> 516 AbstractSNN.set_spiketrain_stats_input(self)
517
518
~/sPyNNaker/lib/python3.8/site-packages/snntoolbox/simulation/utils.py in set_spiketrain_stats_input(self)
1022
1023 if self._poisson_input or self._is_aedat_input:
-> 1024 spiketrains_b_l_t = self.get_spiketrains_input()
1025 if self.input_b_l_t is not None:
1026 self.input_b_l_t = spiketrains_b_l_t
~/sPyNNaker/lib/python3.8/site-packages/snntoolbox/simulation/target_simulators/spiNNaker_target_sim.py in get_spiketrains_input(self)
387 def get_spiketrains_input(self):
388 shape = list(self.parsed_model.input_shape) + [self._num_timesteps]
--> 389 spiketrains_flat = self.layers[0].get_data(
390 'spikes').segments[-1].spiketrains
391 spiketrains_b_l_t = self.reshape_flattened_spiketrains(
~/sPyNNaker/lib/python3.8/site-packages/spynnaker/pyNN/models/populations/population.py in get_data(self, variables, gather, clear, annotations)
376 "not be supported by all platforms.")
377
--> 378 return self._recorder.extract_neo_block(
379 variables, None, clear, annotations)
380
~/sPyNNaker/lib/python3.8/site-packages/spynnaker/pyNN/models/recorder.py in extract_neo_block(self, variables, view_indexes, clear, annotations)
323
324 # add to the segments the new block
--> 325 self.__append_current_segment(block, variables, view_indexes, clear)
326
327 # add fluff to the neo block
~/sPyNNaker/lib/python3.8/site-packages/spynnaker/pyNN/models/recorder.py in __append_current_segment(self, block, variables, view_indexes, clear)
434 self.__read_in_spikes(
435 segment=segment,
--> 436 spikes=self.get_spikes(),
437 t=get_simulator().get_current_time(),
438 n_neurons=self.__population.size,
~/sPyNNaker/lib/python3.8/site-packages/spynnaker/pyNN/models/recorder.py in get_spikes(self)
267 "This population has not got the capability to record spikes")
268 if not self.__vertex.is_recording_spikes():
--> 269 raise ConfigurationException(
270 "This population has not been set to record spikes")
271
ConfigurationException: This population has not been set to record spikes