374 lines
14 KiB
Python
374 lines
14 KiB
Python
|
|
from models.LIFACnoise import LifacNoiseModel
|
|
from CellData import icelldata_of_dir, CellData
|
|
from Baseline import get_baseline_class
|
|
from FiCurve import get_fi_curve_class
|
|
from Fitter import Fitter
|
|
from ModelFit import ModelFit
|
|
|
|
import time
|
|
import os
|
|
import copy
|
|
import argparse
|
|
import numpy as np
|
|
|
|
import multiprocessing as mp
|
|
|
|
|
|
SAVE_PATH_PREFIX = ""
|
|
FIT_ROUTINE = ""
|
|
|
|
|
|
def main():
|
|
parser = argparse.ArgumentParser()
|
|
parser.add_argument("--cell", help="folder (with .dat files) containing the cell data")
|
|
args = parser.parse_args()
|
|
if args.cell is not None:
|
|
cell_data = CellData(args.cell)
|
|
start_parameters = [p for p in iget_start_parameters()]
|
|
fit_cell_parallel(cell_data, start_parameters)
|
|
quit()
|
|
|
|
# test_single_cell("invivo_data/2012-01-17-ap/")
|
|
#
|
|
# quit()
|
|
|
|
start_parameters = [p for p in iget_start_parameters()]
|
|
# start_data = 8
|
|
# count = 0
|
|
# for cell_data in icelldata_of_dir("./invivo_data/"):
|
|
# count += 1
|
|
# if count < start_data:
|
|
# continue
|
|
# fit_cell_parallel(cell_data, start_parameters)
|
|
|
|
cell_data = CellData("invivo_data/2012-04-20-ab-invivo-1/")
|
|
fit_cell_parallel(cell_data, start_parameters)
|
|
|
|
|
|
def test_single_cell(path):
|
|
cell_data = CellData(path)
|
|
start_parameters = [p for p in iget_start_parameters()]
|
|
|
|
for i, p in enumerate(start_parameters):
|
|
fitter = Fitter()
|
|
fitter.set_data_reference_values(cell_data)
|
|
fmin, res_par = fitter.fit_routine_1(p)
|
|
|
|
cell_path = os.path.basename(cell_data.get_data_path())
|
|
|
|
error = fitter.calculate_errors(model=LifacNoiseModel(res_par))
|
|
save_path = "results/invivo_results/" + cell_path + "/start_parameter_{:}_err_{:.2f}/".format(i, sum(error))
|
|
save_fitting_run_info(cell_data, res_par, p, plot=True, save_path=save_path)
|
|
print("Done with start parameters {}".format(str(i)))
|
|
|
|
|
|
def fit_cell_base(parameters):
|
|
# parameter = (cell_data, start_parameter_index, start_parameter, results_base_folder)
|
|
time1 = time.time()
|
|
fitter = Fitter()
|
|
fitter.set_data_reference_values(parameters[0])
|
|
fmin, res_par = fitter.fit_routine_1(parameters[2])
|
|
|
|
cell_data = parameters[0]
|
|
cell_path = os.path.split(cell_data.get_data_path())[-1]
|
|
|
|
error = fitter.calculate_errors(model=LifacNoiseModel(res_par))
|
|
save_path = parameters[3] + "/" + cell_path + "/start_parameter_{:}_err_{:.2f}/".format(parameters[1], sum(error))
|
|
save_fitting_run_info(parameters[0], res_par, parameters[2], plot=True, save_path=save_path)
|
|
time2 = time.time()
|
|
|
|
del fitter
|
|
|
|
print("Time taken for " + cell_path +
|
|
"\n and start parameters ({:}): {:.2f}s thread time".format(parameters[1]+1, time2 - time1) +
|
|
"\n error: {:.2f}".format(sum(error)))
|
|
|
|
|
|
def fit_all_cells_parallel_sync(cells, start_parameters, thread_pool, results_base_folder):
|
|
parameter = []
|
|
for cell in cells:
|
|
for i, s_pars in enumerate(start_parameters):
|
|
parameter.append((cell, i, s_pars, results_base_folder))
|
|
|
|
time1 = time.time()
|
|
thread_pool.map(fit_cell_base, parameter)
|
|
time2 = time.time()
|
|
print("Time taken for all ({:}): {:.2f}s".format(len(parameter)*len(cells), time2 - time1))
|
|
|
|
|
|
def fit_cell_parallel(cell_data, start_parameters):
|
|
cell_path = os.path.basename(cell_data.get_data_path())
|
|
save_directory = "./results/invivo_results/"
|
|
save_path_cell = os.path.join(save_directory, cell_data.get_cell_name())
|
|
print(cell_path)
|
|
core_count = mp.cpu_count()
|
|
pool = mp.Pool(core_count - 1)
|
|
|
|
parameters = []
|
|
for i, p in enumerate(start_parameters):
|
|
parameters.append((cell_data, i, p, save_directory))
|
|
|
|
time1 = time.time()
|
|
pool.map(fit_cell_base, parameters)
|
|
time2 = time.time()
|
|
print("Time taken for all start parameters ({:}): {:.2f}s".format(len(start_parameters), time2-time1))
|
|
del pool
|
|
del cell_data
|
|
|
|
best_fit = None
|
|
min_err = np.inf
|
|
for fit in os.listdir(save_path_cell):
|
|
cur_fit = ModelFit(os.path.join(save_path_cell, fit))
|
|
if cur_fit.comparable_error() < min_err:
|
|
min_err = cur_fit.comparable_error()
|
|
best_fit = cur_fit
|
|
|
|
best_fit.generate_master_plot("./results/invivo_best/singles/")
|
|
|
|
|
|
|
|
def test_fit_routines():
|
|
fitter = Fitter()
|
|
names = ("routine_1", "routine_2")
|
|
global FIT_ROUTINE
|
|
for i, routine in enumerate([fitter.fit_routine_1, fitter.fit_routine_2]):
|
|
FIT_ROUTINE = names[i]
|
|
run_with_real_data(fitter, routine)
|
|
|
|
best = []
|
|
cells = sorted(os.listdir("test_routines/" + names[0] + "/"))
|
|
for name in names:
|
|
|
|
save_path = "test_routines/" + name + "/"
|
|
cell_best = []
|
|
for directory in sorted(os.listdir(save_path)):
|
|
path = os.path.join(save_path, directory)
|
|
if os.path.isdir(path):
|
|
cell_best.append(find_best_run(path))
|
|
best.append(cell_best)
|
|
|
|
with open("test_routines/comparision.csv", "w") as res_file:
|
|
res_file.write("routine")
|
|
for cell in cells:
|
|
res_file.write("," + cell)
|
|
|
|
for i, routine_results in enumerate(best):
|
|
res_file.write(names[i])
|
|
for cell_best in routine_results:
|
|
res_file.write("," + str(cell_best))
|
|
|
|
|
|
def find_best_run(cell_path):
|
|
values = []
|
|
for directory in sorted(os.listdir(cell_path)):
|
|
start_par_path = os.path.join(cell_path, directory)
|
|
if os.path.isdir(start_par_path):
|
|
values.append(float(start_par_path.split("_")[-1]))
|
|
|
|
return min(values)
|
|
|
|
|
|
def iget_start_parameters():
|
|
# mem_tau, input_scaling, noise_strength, dend_tau,
|
|
# expand by tau_a, delta_a ?
|
|
|
|
mem_tau_list = [0.01]
|
|
input_scaling_list = [60]
|
|
noise_strength_list = [0.03] # [0.02, 0.06]
|
|
dend_tau_list = [0.001, 0.002]
|
|
delta_a_list = [0.035, 0.065]
|
|
tau_a_list = [0.1, 0.4]
|
|
ref_time_list = [0.00065]
|
|
|
|
for mem_tau in mem_tau_list:
|
|
for input_scaling in input_scaling_list:
|
|
for noise_strength in noise_strength_list:
|
|
for dend_tau in dend_tau_list:
|
|
for delta_a in delta_a_list:
|
|
for tau_a in tau_a_list:
|
|
for ref_time in ref_time_list:
|
|
yield {"mem_tau": mem_tau, "input_scaling": input_scaling,
|
|
"noise_strength": noise_strength, "dend_tau": dend_tau,
|
|
"delta_a": delta_a, "tau_a": tau_a, "refractory_period": ref_time}
|
|
|
|
|
|
def run_with_real_data(fitter, fit_routine_func, parallel=False):
|
|
count = 0
|
|
for cell_data in icelldata_of_dir("./data/"):
|
|
count += 1
|
|
if count < 7:
|
|
pass
|
|
#continue
|
|
|
|
print("cell:", cell_data.get_data_path())
|
|
trace = cell_data.get_base_traces(trace_type=cell_data.V1)
|
|
if len(trace) == 0:
|
|
print("NO V1 TRACE FOUND")
|
|
continue
|
|
|
|
global FIT_ROUTINE
|
|
# results_path = "results/" + os.path.split(cell_data.get_data_path())[-1] + "/"
|
|
results_path = "test_routines/" + FIT_ROUTINE + "/" + os.path.split(cell_data.get_data_path())[-1] + "/"
|
|
print("results at:", results_path)
|
|
|
|
if not os.path.exists(results_path):
|
|
os.makedirs(results_path)
|
|
|
|
# plot cell images:
|
|
cell_save_path = results_path + "cell/"
|
|
if not os.path.exists(cell_save_path):
|
|
os.makedirs(cell_save_path)
|
|
data_baseline = get_baseline_class(cell_data)
|
|
data_baseline.plot_baseline(cell_save_path)
|
|
data_baseline.plot_interspike_interval_histogram(cell_save_path)
|
|
data_baseline.plot_serial_correlation(6, cell_save_path)
|
|
|
|
data_fi_curve = get_fi_curve_class(cell_data, cell_data.get_fi_contrasts())
|
|
data_fi_curve.plot_fi_curve(cell_save_path)
|
|
|
|
start_par_count = 0
|
|
for start_parameters in iget_start_parameters():
|
|
start_par_count += 1
|
|
print("START PARAMETERS:", start_par_count)
|
|
|
|
start_time = time.time()
|
|
# fitter = Fitter()
|
|
fmin, parameters = fitter.fit_model_to_data(cell_data, start_parameters, fit_routine_func)
|
|
|
|
print(fmin)
|
|
print(parameters)
|
|
end_time = time.time()
|
|
parameter_set_path = results_path + "start_par_set_{}_fmin_{:.2f}".format(start_par_count, fmin["fun"]) + "/"
|
|
|
|
print('Fitting of cell took function took {:.3f} s'.format((end_time - start_time)))
|
|
# print(results_path)
|
|
save_fitting_run_info(cell_data, parameters, start_parameters,
|
|
plot=True, save_path=parameter_set_path)
|
|
|
|
# from Sounds import play_finished_sound
|
|
# play_finished_sound()
|
|
pass
|
|
|
|
|
|
def save_fitting_run_info(cell_data, parameters, start_parameters, plot=False, save_path=None):
|
|
if save_path is not None:
|
|
if not os.path.exists(save_path):
|
|
os.makedirs(save_path)
|
|
|
|
if save_path is None:
|
|
return
|
|
|
|
with open(save_path + "parameters_info.txt", "w") as file:
|
|
file.writelines(["start_parameters:\t" + str(start_parameters),
|
|
"\nfinal_parameters:\t" + str(parameters)])
|
|
|
|
model = LifacNoiseModel(parameters)
|
|
eod_frequency = cell_data.get_eod_frequency()
|
|
|
|
data_baseline = get_baseline_class(cell_data)
|
|
c_bf = data_baseline.get_baseline_frequency()
|
|
c_vs = data_baseline.get_vector_strength()
|
|
c_sc = data_baseline.get_serial_correlation(1)
|
|
c_cv = data_baseline.get_coefficient_of_variation()
|
|
c_burst = data_baseline.get_burstiness()
|
|
|
|
data_fi_curve = get_fi_curve_class(cell_data, cell_data.get_fi_contrasts())
|
|
c_f_inf_slope = data_fi_curve.get_f_inf_slope()
|
|
c_f_inf_values = data_fi_curve.f_inf_frequencies
|
|
|
|
c_f_zero_slope = data_fi_curve.get_f_zero_fit_slope_at_straight()
|
|
c_f_zero_values = data_fi_curve.f_zero_frequencies
|
|
|
|
if c_f_inf_slope < 0:
|
|
contrasts = np.array(cell_data.get_fi_contrasts()) * -1
|
|
else:
|
|
contrasts = np.array(cell_data.get_fi_contrasts())
|
|
model_baseline = get_baseline_class(model, eod_frequency, trials=15)
|
|
m_bf = model_baseline.get_baseline_frequency()
|
|
m_vs = model_baseline.get_vector_strength()
|
|
m_sc = model_baseline.get_serial_correlation(1)
|
|
m_cv = model_baseline.get_coefficient_of_variation()
|
|
m_burst = model_baseline.get_burstiness()
|
|
|
|
model_ficurve = get_fi_curve_class(model, cell_data.get_fi_contrasts(), eod_frequency, trials=60)
|
|
m_f_infinities = model_ficurve.get_f_inf_frequencies()
|
|
m_f_zeros = model_ficurve.get_f_zero_frequencies()
|
|
m_f_infinities_slope = model_ficurve.get_f_inf_slope()
|
|
m_f_zero_slope = model_ficurve.get_f_zero_fit_slope_at_straight()
|
|
|
|
np.save(os.path.join(save_path, "model_fi_inf_values.npy"), np.array(m_f_infinities))
|
|
np.save(os.path.join(save_path, "cell_fi_inf_values.npy"), np.array(c_f_inf_values))
|
|
np.save(os.path.join(save_path, "model_fi_zero_values.npy"), np.array(m_f_zeros))
|
|
np.save(os.path.join(save_path, "cell_fi_zero_values.npy"), np.array(c_f_zero_values))
|
|
|
|
with open(os.path.join(save_path, "cell_data_path.txt"), "w") as f:
|
|
path = cell_data.get_data_path() + "\n"
|
|
f.write(path)
|
|
|
|
if c_f_inf_slope < 0:
|
|
model_ficurve.stimulus_values = contrasts * -1
|
|
|
|
# print("EOD-frequency: {:.2f}".format(cell_data.get_eod_frequency()))
|
|
# print("bf: cell - {:.2f} vs model {:.2f}".format(c_bf, m_bf))
|
|
# print("vs: cell - {:.2f} vs model {:.2f}".format(c_vs, m_vs))
|
|
# print("sc: cell - {:.2f} vs model {:.2f}".format(c_sc[0], m_sc[0]))
|
|
# print("cv: cell - {:.2f} vs model {:.2f}".format(c_cv, m_cv))
|
|
# print("f_inf_slope: cell - {:.2f} vs model {:.2f}".format(c_f_inf_slope, m_f_infinities_slope))
|
|
# print("f infinity values:\n cell -", c_f_inf_values, "\n model -", m_f_infinities)
|
|
#
|
|
# print("f_zero_slope: cell - {:.2f} vs model {:.2f}".format(c_f_zero_slope, m_f_zero_slope))
|
|
# print("f zero values:\n cell -", c_f_zero_values, "\n model -", m_f_zeros)
|
|
if save_path is not None:
|
|
with open(save_path + "value_comparision.tsv", 'w') as value_file:
|
|
value_file.write("Variable\tCell\tModel\n")
|
|
value_file.write("baseline_frequency\t{:.2f}\t{:.2f}\n".format(c_bf, m_bf))
|
|
value_file.write("vector_strength\t{:.2f}\t{:.2f}\n".format(c_vs, m_vs))
|
|
value_file.write("serial_correlation\t{:.2f}\t{:.2f}\n".format(c_sc[0], m_sc[0]))
|
|
value_file.write("Burstiness\t{:.2f}\t{:.2f}\n".format(c_burst, m_burst))
|
|
value_file.write("coefficient_of_variation\t{:.2f}\t{:.2f}\n".format(c_cv, m_cv))
|
|
value_file.write("f_inf_slope\t{:.2f}\t{:.2f}\n".format(c_f_inf_slope, m_f_infinities_slope))
|
|
value_file.write("f_zero_slope\t{:.2f}\t{:.2f}\n".format(c_f_zero_slope, m_f_zero_slope))
|
|
|
|
if plot:
|
|
# plot model images
|
|
model_baseline.plot_baseline(save_path)
|
|
model_baseline.plot_interspike_interval_histogram(save_path)
|
|
model_baseline.plot_isi_histogram_comparision(data_baseline.get_interspike_intervals(),
|
|
model_baseline.get_interspike_intervals(), save_path)
|
|
model_baseline.plot_serial_correlation(6, save_path)
|
|
|
|
model_ficurve.plot_fi_curve(save_path)
|
|
model_ficurve.plot_fi_curve_comparision(data_fi_curve, model_ficurve, save_path)
|
|
|
|
|
|
def test_effect_of_refractory_period():
|
|
ref_periods = [0.0006, 0.001, 0.0015]
|
|
counter = 0
|
|
core_count = mp.cpu_count()
|
|
|
|
for cell in icelldata_of_dir("./data/"):
|
|
pool = mp.Pool(core_count - 1)
|
|
counter += 1
|
|
if counter < 10:
|
|
continue
|
|
elif counter >= 14:
|
|
return
|
|
start_parameters_base = [p for p in iget_start_parameters()]
|
|
for ref_period in ref_periods:
|
|
print(cell.get_data_path())
|
|
print("ref period: {:.4f}".format(ref_period))
|
|
results_base_folder = "./test_routines/ref_period_{:.4f}/".format(ref_period)
|
|
all_start_parameters = copy.deepcopy(start_parameters_base)
|
|
|
|
for par_set in all_start_parameters:
|
|
par_set["refractory_period"] = ref_period
|
|
fit_all_cells_parallel_sync([cell], all_start_parameters, pool, results_base_folder)
|
|
|
|
del cell
|
|
del pool
|
|
|
|
|
|
if __name__ == '__main__':
|
|
main()
|