Fix inconsistency of `scale_down_to`

6d883048 · Johanna Senk · GitHub · b5fa5c99 · 6d883048
Unverified Commit 6d883048 authored 2 years ago by Johanna Senk Committed by GitHub 2 years ago
--- a/multi-area-model.ipynb
+++ b/multi-area-model.ipynb
@@ -71,7 +71,7 @@
   "id": "2cedd26b",
   "metadata": {},
   "source": [
-    "Neurons and indegrees are both scaled down to 1%.\n",
+    "Neurons and indegrees are both scaled down to 0.5%.\n",
    "Can usually be simulated on a local machine.\n",
    "\n",
    "**Warning: This will not yield reasonable dynamical results from the\n",

 %% Cell type:markdown id:b1331599 tags:
 # Down-scaled multi-area model
 %% Cell type:code id:96517739 tags:
 ``` python
 %matplotlib inline
 import matplotlib.pyplot as plt
 import numpy as np
 import os
 ```
 %% Cell type:code id:7e07b0d0 tags:
 ``` python
 !pip install nested_dict dicthash
 ```
 %% Cell type:markdown id:0c6b6a3c tags:
 Create config file.
 %% Cell type:code id:72c170e4 tags:
 ``` python
 with open('config.py', 'w') as fp:
    fp.write(
 '''import os
 base_path = os.path.abspath(".")
 data_path = os.path.abspath("simulations")
 jobscript_template = "python {base_path}/run_simulation.py {label}"
 submit_cmd = "bash -c"
 ''')
 ```
 %% Cell type:code id:2784f76b tags:
 ``` python
 from multiarea_model import MultiAreaModel
 ```
 %% Cell type:markdown id:2cedd26b tags:
-Neurons and indegrees are both scaled down to 1%.
+Neurons and indegrees are both scaled down to 0.5%.
 Can usually be simulated on a local machine.
 **Warning: This will not yield reasonable dynamical results from the
 network and is only meant to demonstrate the simulation workflow.**
 %% Cell type:code id:e940bb6b tags:
 ``` python
 scale_down_to = 0.005
 ```
 %% Cell type:markdown id:d53f1eab tags:
 Specify model and simulation parameters.
 %% Cell type:code id:7af3a191 tags:
 ``` python
 conn_params = {'replace_non_simulated_areas': 'het_poisson_stat',
               'g': -11.,
               'K_stable': 'K_stable.npy',
               'fac_nu_ext_TH': 1.2,
               'fac_nu_ext_5E': 1.125,
               'fac_nu_ext_6E': 1.41666667,
               'av_indegree_V1': 3950.}
 input_params = {'rate_ext': 10.}
 neuron_params = {'V0_mean': -150.,
                 'V0_sd': 50.}
 network_params = {'N_scaling': scale_down_to,
                  'K_scaling': scale_down_to,
                  'fullscale_rates': 'tests/fullscale_rates.json',
                  'input_params': input_params,
                  'connection_params': conn_params,
                  'neuron_params': neuron_params}
 sim_params = {'t_sim': 2000.,
              'num_processes': 1,
              'local_num_threads': 1,
              'recording_dict': {'record_vm': False}}
 theory_params = {'dt': 0.1}
 ```
 %% Cell type:markdown id:de4a6703 tags:
 Instantiate a multi-area model, predict firing rates from theroy, and run the simulation.
 %% Cell type:code id:d409be95 tags:
 ``` python
 M = MultiAreaModel(network_params, simulation=True,
                   sim_spec=sim_params,
                   theory=True,
                   theory_spec=theory_params)
 ```
 %% Cell type:code id:918d907f tags:
 ``` python
 p, r = M.theory.integrate_siegert()
 print("Mean-field theory predicts an average "
      "rate of {0:.3f} spikes/s across all populations.".format(np.mean(r[:, -1])))
 ```
 %% Cell type:code id:15778e9c tags:
 ``` python
 M.simulation.simulate()
 ```
 %% Cell type:markdown id:8726a93d tags:
 Load spike data.
 %% Cell type:code id:cb8e3edd tags:
 ``` python
 data = np.loadtxt(M.simulation.data_dir + '/recordings/' + M.simulation.label + "-spikes-1-0.dat", skiprows=3)
 ```
 %% Cell type:markdown id:8793e033 tags:
 Compute instantaneous rate per neuron across all populations.
 %% Cell type:code id:9590223b tags:
 ``` python
 tsteps, spikecount = np.unique(data[:,1], return_counts=True)
 rate = spikecount / M.simulation.params['dt'] * 1e3 / np.sum(M.N_vec)
 ```
 %% Cell type:markdown id:38ddd973 tags:
 Plot instantaneous and mean rate.
 %% Cell type:code id:bea30fc8 tags:
 ``` python
 fig, ax = plt.subplots()
 ax.plot(tsteps, rate)
 ax.plot(tsteps, np.average(rate)*np.ones(len(tsteps)), label='mean')
 ax.set_title('instantaneous rate across all populations')
 ax.set_xlabel('time (ms)')
 ax.set_ylabel('rate (spikes / s)')
 ax.set_xlim(0, sim_params['t_sim'])
 ax.set_ylim(0, 50)
 ax.legend()
 ```