import csv

import neuron

from neuron import h

from neuron import gui

import matplotlib.pyplot as plt

import subprocess

from neuron import load_mechanisms

import random

import math

import numpy as np

import itertools

# just adding some comments

class cell(object):

self.soma.gkbar_hh=.038

def set_stim(seg,amp,delay,dur):

return stim

def recvolt(seg):

rec_v=neuron.h.Vector()

rec_v.record(seg._ref_v)

return rec_v

def rectime(seg):

rec_t=neuron.h.Vector()

rec_t.record(seg._ref_t)

return rec_t

load_mechanisms('/home/ben/Desktop/5ht3a_project/neuron_model/downloaded_neuron_models/Hipp_paper_code/')

h.xopen('/home/ben/Desktop/5ht3a_project/neuron_model/downloaded_neuron_models/Hipp_paper_code/basket_cell17S.hoc')

h.xopen('/home/ben/Desktop/5ht3a_project/neuron_model/downloaded_neuron_models/Hipp_paper_code/pyramidal_cell_14Vb.hoc')

h.xopen('/home/ben/Desktop/5ht3a_project/neuron_model/downloaded_neuron_models/Hipp_paper_code/olm_cell2.hoc')

pyrstren=np.arange(.005,.01,.001)

pvstren=np.arange(0,0.3,.05)

olmexcstren=np.arange(.05,.055,.001)

olmstren=np.arange(.04,.05,.002)

olmcombs=itertools.product(olmexcstren,olmstren)

pvpyrcombs=itertools.product(pyrstren,pvstren)

numiter=100

ppp=1

for ggg in olmcombs:

print 'run#'+str(ppp)+'/30'

ppp=ppp+1

spikecount=np.zeros((numiter,3))

for hhh in range(0,numiter):

sccell=cell()

rec_v_sccell=recvolt(sccell.soma(0.5))

sccell_stim=set_stim(sccell.soma,1,100,2)

rec_v_sccell=recvolt(sccell.soma(0.5))

pyrcell=h.PyramidalCell()

pyrcell.soma.Ra=30

#pyrstim=set_stim(pyrcell.soma,2)

rec_v_pyrcell=recvolt(pyrcell.soma(0.5))

synpyr=neuron.h.Exp2Syn(pyrcell.radTprox(0))

synpyr.tau1=0.5

synpyr.tau2=3

synpyr.e=0

pvin=h.BasketCell()

pvin.soma.Ra=200

pvin.soma.cm=1

#pvstim=set_stim(pvin.soma,2)

rec_v_pvin=recvolt(pvin.soma(0.5))

synpv=neuron.h.Exp2Syn(pvin.radT2(0))

synpvpyr=neuron.h.Exp2Syn(pyrcell.soma(0.5))

synpv.tau1=0.5

synpv.tau2=3

synpv.e=0

synpvpyr.tau1=0.5

synpvpyr.tau2=3

synpvpyr.e=-75

###SPECIFY THE VALUES OF THE SECTION TO BE RECORDED##

#record time from NEURON (neuron.h._ref_t)

rec_t=h.Vector()

rec_t.record(neuron.h._ref_t)

spontpyrepsc=neuron.h.Exp2Syn(pyrcell.soma(0.5))

spontpyrepsc.tau1=0.5

spontpyrepsc.tau2=3

spontpyrepsc.e=0

spontpyripsc=neuron.h.Exp2Syn(pyrcell.soma(0.5))

spontpyripsc.tau1=1

spontpyripsc.tau2=8

spontpyripsc.e=-75

#######Inititate OLM cell and define synapses##############

olmcell=h.OLMCell()

rec_v_olm=recvolt(olmcell.soma(0.5))

#olmstim=set_stim(olmcell.soma,.1)

synolm=neuron.h.Exp2Syn(olmcell.dend1(0))

synolm.tau1=0.5

synolm.tau2=3

synolm.e=0

synolmpv=neuron.h.Exp2Syn(pvin.soma(0.5))

synolmpv.tau1=2

synolmpv.tau2=10

synolmpv.e=-75

synolmpyr=neuron.h.Exp2Syn(pyrcell.soma(0.5))

synolmpyr.tau1=2

synolmpyr.tau2=10

synolmpyr.e=-75

spontolmepsc=neuron.h.Exp2Syn(olmcell.soma(0.5))

spontolmepsc.tau1=0.5

spontolmepsc.tau2=3

spontolmepsc.e=0

spontolmipsc=neuron.h.Exp2Syn(olmcell.soma(0.5))

spontolmipsc.tau1=1

spontolmipsc.tau2=8

spontolmipsc.e=-75

#############################################################

#####################synaptic connections########################################

####lag between EPSP and IPSP should be ~3 ms from Karnup and Stelzer (1999, J. Physiol)

ampasyn=neuron.h.NetCon(sccell.soma(0.5)._ref_v,synpyr,0,0,.008,sec=sccell.soma)#original conductance=.006; somewhere between .0042 and .0045 pyr cells fire 100%

#probability of exciting pyr with no inhibition: .0044=.94, .0045=.98, .0046=.97,.0047=.97,.005=.95,.006=.98,.0065=.99,.007=1,.01=1

ampasynpv=neuron.h.NetCon(sccell.soma(0.5)._ref_v,synpv,0,0,0.04,sec=sccell.soma)#original weight=.03

gabasynpvpyr=neuron.h.NetCon(pvin.soma(0.5)._ref_v,synpvpyr,-40,1,.2,sec=pvin.soma) #original weight=.01

ampasynolm=neuron.h.NetCon(sccell.soma(0.5)._ref_v,synolm,-20,0,ggg[0],sec=sccell.soma)#original weight=.01

gabasynolmpv=neuron.h.NetCon(olmcell.soma(0.5)._ref_v,synolmpv,-30,0,ggg[1],sec=olmcell.soma)#original weight=.04

gabasynolmpyr=neuron.h.NetCon(olmcell.soma(0.5)._ref_v,synolmpyr,-30,0,0,sec=olmcell.soma)#original weight=.01

###############################################################

###spontaneous events on pyramidal cell######

#generate poisson process giving sepsc and sipsc times in time course

rateconst = float(2000)

timekeep=0

events=[]

eventtimespyrepsc=[]

for i in range(1,14003):

ppval=1.0-math.exp(-(i-timekeep)/rateconst)

if ppval>random.random():

events.append(1)

eventtimespyrepsc.append(i)

timekeep=i

else:

events.append(0)

timekeep=i

spontpyrepsc=list()

for jj in range(0,len(eventtimespyrepsc)):

spontpyrepsc.append(neuron.h.AlphaSynapse(pyrcell.soma(0.5)))

spontpyrepsc[jj].onset=(eventtimespyrepsc[jj]/14002.0)*350.0

#print "eventimes[jj]="+str(spontpyrepsc[jj].onset)

spontpyrepsc[jj].e=0

spontpyrepsc[jj].gmax=.002*np.random.lognormal(.002,.2,1)

spontpyrepsc[jj].tau=1

timekeep=0

events=[]

eventtimespyripsc=[]

for i in range(1,14003):

ppval=1.0-math.exp(-(i-timekeep)/rateconst)

if ppval>random.random():

events.append(1)

eventtimespyripsc.append(i)

timekeep=i

else:

events.append(0)

timekeep=i

spontpyripsc=list()

for jj in range(0,len(eventtimespyripsc)):

spontpyripsc.append(neuron.h.AlphaSynapse(pyrcell.soma(0.5)))

spontpyripsc[jj].onset=(eventtimespyripsc[jj]/14002.0)*350.0

#print "eventimes[jj]="+str(spontpyripsc[jj].onset)

spontpyripsc[jj].e=-75

spontpyripsc[jj].gmax=.002*np.random.lognormal(.002,.2,1)

spontpyripsc[jj].tau=5

######################################################

####Spontaneous synaptic events on PV cell#############

timekeep=0

events=[]

eventtimespvepsc=[]

for i in range(1,14003):

ppval=1.0-math.exp(-(i-timekeep)/rateconst)

if ppval>random.random():

events.append(1)

eventtimespvepsc.append(i)

timekeep=i

else:

events.append(0)

timekeep=i

spontpvepsc=list()

for jj in range(0,len(eventtimespvepsc)):

spontpvepsc.append(neuron.h.AlphaSynapse(pvin.soma(0.5)))

spontpvepsc[jj].onset=(eventtimespvepsc[jj]/14002.0)*350.0

#print "eventimes[jj]="+str(spontpvepsc[jj].onset)

spontpvepsc[jj].e=0

spontpvepsc[jj].gmax=.002*np.random.lognormal(.002,.2,1)

spontpvepsc[jj].tau=1

rateconst=float(2000)

timekeep=0

events=[]

eventtimespvipsc=[]

for i in range(1,14003):

ppval=1.0-math.exp(-(i-timekeep)/rateconst)

if ppval>random.random():

events.append(1)

eventtimespvipsc.append(i)

timekeep=i

else:

events.append(0)

timekeep=i

spontpvipsc=list()

for jj in range(0,len(eventtimespvipsc)):

spontpvipsc.append(neuron.h.AlphaSynapse(pvin.soma(0.5)))

spontpvipsc[jj].onset=(eventtimespvipsc[jj]/14002.0)*350.0

#print "eventimes[jj]="+str(spontpvipsc[jj].onset)

spontpvipsc[jj].e=-75

spontpvipsc[jj].gmax=.002*np.random.lognormal(.002,.2,1)

spontpvipsc[jj].tau=5

######################################################

####Spontaneous synaptic events on OLM cell#############

timekeep=0

events=[]

eventtimesolmepsc=[]

for i in range(1,14003):

ppval=1.0-math.exp(-(i-timekeep)/rateconst)

if ppval>random.random():

events.append(1)

eventtimesolmepsc.append(i)

timekeep=i

else:

events.append(0)

timekeep=i

spontolmepsc=list()

for jj in range(0,len(eventtimesolmepsc)):

spontolmepsc.append(neuron.h.AlphaSynapse(olmcell.soma(0.5)))

spontolmepsc[jj].onset=(eventtimesolmepsc[jj]/14002.0)*350.0

#print "eventimes[jj]="+str(spontolmepsc[jj].onset)

spontolmepsc[jj].e=0

spontolmepsc[jj].gmax=.002*np.random.lognormal(.002,.2,1)

spontolmepsc[jj].tau=1

rateconst=float(2000)

timekeep=0

events=[]

eventtimesolmipsc=[]

for i in range(1,14003):

ppval=1.0-math.exp(-(i-timekeep)/rateconst)

if ppval>random.random():

events.append(1)

eventtimesolmipsc.append(i)

timekeep=i

else:

events.append(0)

timekeep=i

spontolmipsc=list()

for jj in range(0,len(eventtimesolmipsc)):

spontolmipsc.append(neuron.h.AlphaSynapse(olmcell.soma(0.5)))

spontolmipsc[jj].onset=(eventtimesolmipsc[jj]/14002.0)*350.0

#print "eventimes[jj]="+str(spontolmipsc[jj].onset)

spontolmipsc[jj].e=-75

spontolmipsc[jj].gmax=.002*np.random.lognormal(.002,.2,1)

spontolmipsc[jj].tau=5

######################################################

######constant currents############

#set_stim(seg,amp,delay,duration)

pyrstim=set_stim(pyrcell.soma,0,0,0)

pvstm=set_stim(pvin.soma,0,0,0)

olmcellstim=set_stim(olmcell.soma,-.5,0,350)

#################################

h.finitialize(-65)

#set time of the simulation

tstop=350

#Run the simulation

neuron.run(tstop)

time=rec_t.as_numpy()

pvin_volt=rec_v_pvin.as_numpy()

pyrcell_volt=rec_v_pyrcell.as_numpy()

olmcell_volt=rec_v_olm.as_numpy()

sccell_volt=rec_v_sccell.as_numpy()

'''

plt.plot(time,olmcell_volt,color='b',linewidth=3.0)

plt.show()

print "length pvin_volt="+str(len(pvin_volt))

plt.plot(time,pvin_volt,color='r',linewidth=3.0)

#plt.plot(time,olmcell_volt,color='b',linewidth=3.0)

plt.plot(time,sccell_volt,color='g',linewidth=3.0)

plt.plot(time,pyrcell_volt,color='k',linewidth=3.0)

plt.show()

'''

'''

print "pv Rm="+str(pvin.soma.Ra)

print "pyr Rm="+str(pyrcell.soma.Ra)

print "length of v vectors:"+str(len(pyrcell_volt))

print "mean of events="+str(np.mean(events))

print "length of events="+str(len(events))

print "length of spont epsc="+str(len(spontpyrepsc))

'''

'''

print "max value pyr:"+str(max(pyrcell_volt[4081:4400]))#search for max in (4081:4400)*350/14002=102~110 ms

print "max value pv:"+str(max(pvin_volt[4081:4400]))

print "max value pv:"+str(max(olmcell_volt[4081:4400]))

print "iteration:"+str(hhh+1)

'''

if max(pyrcell_volt[4081:4400])>-20:

spikecount[hhh,0]=1

if max(pvin_volt[4081:4400])>0:

spikecount[hhh,1]=1

if max(olmcell_volt[4081:4400])>0:

spikecount[hhh,2]=1

'''

print "pyrstren="+str(ggg[0])

print "pvstren="+str(ggg[1])

'''

print "mean pyr:"+str(np.mean(spikecount[:,0]))

with open('test2.csv','a') as testfile:

csvwriter=csv.writer(testfile)

csvwriter.writerow([ggg[0],ggg[1],np.mean(spikecount[:,0])])

print "mean pv:"+str(np.mean(spikecount[:,1]))

print "mean olm:"+str(np.mean(spikecount[:,2]))