print('i = %d'%(i))
data = np.load(foldername + '/Series%s'%(k_id) + '/param.npy', allow_pickle=True)
param = data.item()
for key, val in param.items():
exec(key + '=val')
# Import model simulation function and stimulus generation function.
sys.path.append('%s/Parameters'%(codedirectory))
from model_simulation_all import simulation, stimulate
T_all = T + T_relax # simulation time and randomization of initial condition
spike_time_duration = []
for k in range(rep):
print('Loop at %d\n' %(k))
seednumber = i*100+k # i is Series number. k is repitition number.
va = simulation(model, tau, posNa, T_all, mean, std, dt=dt, seednumber=seednumber, stim_type=stim_type, amplitude=amplitude, frequency=frequency)
a = np.diff((np.array(va)>spthr)*1.0)
sp_id = np.where(a==1)[0] # spike time index
b = np.diff((np.array(va)>v_loc)*1.0)
v_loc_id_all = np.where(b==1)[0] # index of passing v_loc from below
v_loc_pair = [] # the last time to pass v_loc from below before each spike time
for sp in sp_id:
if sum(np.diff((v_loc_id_all<sp)*1.0)) == 0: # if sp is larger than all the v_loc_id, then the last one is paired.
v_loc_pair.append(v_loc_id_all[-1])
else:
c = np.diff((v_loc_id_all<sp)*1.0)
v_loc_last_id = np.where(c==-1)[0][-1]
v_loc_pair.append(v_loc_id_all[v_loc_last_id])
spike_time_duration += list((np.array(list(sp_id)) - np.array(v_loc_pair))*dt)
sys.stdout.flush()
#! /usr/bin/python
'''
Firing rate estimation
'''
import sys, os
import numpy as np
foldername = sys.argv[-1]
# Parameters loading
data = np.load(foldername + '/param.npy', allow_pickle=True)
param = data.item()
for key, val in param.items():
exec(key + '=val')
os.environ.keys()
i = int(os.environ['SLURM_ARRAY_TASK_ID'])
stim_mean = Iconst[i-1]
from model_simulation_all import simulation
va = simulation(model, 5, posNa, T, stim_mean, 0, dt=dt) # stim_std = 0
a = np.diff((np.array(va)>spthr)*1)
itemindex = np.where(a==1)
T_s = T/1000.0
fr = len(itemindex[0])/T_s
print("Firing rate is %s Hz"%(fr))
np.save('%s/mean_fr_%i'%(foldername, i),{'mean':stim_mean, 'fr':fr})