def easy_test():
# instanciate model
model = gLIF_model()
# load data
(file_id_list, X_list, Y_list,
dt) = load_neuron_data(1, input_type="noise_only", max_file_ct=4)
# fit model
model.fit(X_list, Y_list, dt)
# predict data via model
Y_pred_list = model.predict(X_list, dt)
# plot predicted data vs actual data for the first voltage trace in X_list
t_arr = dt * np.arange(len(Y_pred_list[0]))
pylab.plot(t_arr, Y_list[0], color='blue', label='True voltage')
pylab.plot(t_arr, Y_pred_list[0], color='green', label='Predicted voltage')
pylab.legend()
pylab.xlabel("Time (s)")
pylab.ylabel("Voltage (mV)")
pylab.title("Response to Input Current")
pylab.show()
def easy_test():
# instanciate model
model = gLIF_model()
# load data
(file_id_list,X_list,Y_list,dt) = load_neuron_data(1,input_type="noise_only",max_file_ct=4)
# fit model
model.fit(X_list, Y_list, dt)
# predict data via model
Y_pred_list = model.predict(X_list, dt)
# plot predicted data vs actual data for the first voltage trace in X_list
t_arr = dt * np.arange(len(Y_pred_list[0]))
pylab.plot(t_arr,Y_list[0],color='blue',label='True voltage')
pylab.plot(t_arr,Y_pred_list[0],color='green',label='Predicted voltage')
pylab.legend()
pylab.xlabel("Time (s)")
pylab.ylabel("Voltage (mV)")
pylab.title("Response to Input Current")
pylab.show()
def test_eval_sav_results(output_dir=OUTPUT_DIR,neuron_num=1):
"""
Example procedure for evaluating the estimated model parameters
gerenated by :func:`test_fit_neuron`.
:param output_dir: directory where JSON files and pickle files summarizing fitting results have be saved.
:param neuron_num: number of neuron we want to evaluate (number ranges from 1 to 12)
.. warning::
This function will not work unless :func:`test_fit_neuron` has
already been run before.
"""
# ------------ LOADING SAVED RESULTS -----------------
folder_id = "neuron_" + str(neuron_num)
neuron_folder = os.path.join(output_dir,folder_id)
if not os.path.exists(neuron_folder):
raise ValueError("Neuron data cannot be found in " + str(neuron_folder))
pickle_file_dir = os.path.join(neuron_folder,"pickle_files")
pickle_file_path = os.path.join(pickle_file_dir,folder_id + ".p")
neuron = pickle.load(open(pickle_file_path,'rb'))
(file_id_list,input_current_list,membrane_voltage_list,dt) = load_neuron_data(neuron_num,
input_type="noise_only",
max_file_ct=4)
# -------------- PLOTTING OUTPUT FIGURES --------------------
print "Running monte carlo simulations..."
simulated_voltage_dict = evaluate.simulate(neuron=neuron,
input_current_list=input_current_list,
membrane_voltage_list=membrane_voltage_list,
file_id_list=file_id_list,
reps=10)
bio_voltage_dict = dict(zip(file_id_list,membrane_voltage_list))
input_current_dict = dict(zip(file_id_list,input_current_list))
figure_dir = os.path.join(neuron_folder,"figures")
if not os.path.exists(figure_dir):
os.makedirs(figure_dir)
# plots simulation results in a specified directory
evaluate.plot_sim_vs_real(simulated_voltage_dict=simulated_voltage_dict,
bio_voltage_dict=bio_voltage_dict,
input_current_dict=input_current_dict,
fig_dir=figure_dir)
stats_dir = os.path.join(neuron_folder,"stats")
if not os.path.exists(stats_dir):
os.makedirs(stats_dir)
evaluate.plot_spk_performance_metrics(bio_voltage_dict,simulated_voltage_dict,fig_dir=stats_dir)
def test_fit_neuron(output_dir=OUTPUT_DIR,neuron_num = 1):
"""
Example procedure for estimating a LIF model for a single neuron.
:param output_dir: directory where JSON files and pickle files summarizing fitting results will be saved.
:param neuron_num: number of neuron we want to fit (number ranges from 1 to 12)
.. note::
The output JSON files will contain a dictionary of parameter values.
The output pickle file will contain a :mod:`fit_neuron.optimize.neuron_base_obj.Neuron` instance
which can simulate the artificial neuron.
"""
if not os.path.exists(output_dir):
os.makedirs(output_dir)
# -------------- LOADING DATA -----------------
# if my_input_type is set to "noise_only" then only load noisy inputs
(file_id_list,input_current_list,membrane_voltage_list,dt) = load_neuron_data(neuron_num,
input_type="noise_only",
max_file_ct=4)
# -------------- RUNNING OPTIMIZATION -----------------
t_bins = T_BIN_DEFAULT
sic_list = [sic_lib.StepSic(t,dt=dt) for t in t_bins]
neuron = fit_neuron.optimize.fit_neuron(input_current_list=input_current_list,
membrane_voltage_list=membrane_voltage_list,
dt=dt,
process_ct=None,
max_lik_iter_max=25,
stopping_criteria=0.01,
sic_list=sic_list)
# -------------- SAVING RESULTS -----------------
folder_id = "neuron_" + str(neuron_num)
neuron_folder = os.path.join(output_dir,folder_id)
if not os.path.exists(neuron_folder):
os.makedirs(neuron_folder)
pickle_file_dir = os.path.join(neuron_folder,"pickle_files")
if not os.path.exists(pickle_file_dir):
os.makedirs(pickle_file_dir)
pickle_file_path = os.path.join(pickle_file_dir,folder_id + ".p")
print "Successfully pickle file to: " + str(pickle_file_path)
pickle.dump(neuron, open(pickle_file_path,'wb'))
json_file_dir = os.path.join(neuron_folder,"json_files")
if not os.path.exists(json_file_dir):
os.makedirs(json_file_dir)
json_file_path = os.path.join(json_file_dir,folder_id + ".json")
json.dump(neuron.get_param_dict(),open(json_file_path,'wb'),sort_keys = False, indent = 4)
print "Successfully saved json file to: " + str(json_file_path)
