def runonce(x, P):
####GENERATE SPIKE INPUTS (the spiketimes are always the same to improve comparability)
thisdur = (P["dur"][x] - 5.0) / 1000.0
N = P["Nsyn"][x]
anf_num = (N + 1, 0, 0)
SO = seh.create_sound(
fs=100e3, freq=P["freq"][x], duration=thisdur, dbspl=P["dB"][x]
)
SP = seh.create_spk(
SO, fs=100e3, N=P["Nsyn"][x] + 1, cf=P["cf"][x], seed=65564 + x, anf_num=anf_num
) # generate enough spiketrains
S = np.array(SP["spikes"] * 1000.0)
# somaticsynapsesonly=0,
R0 = sem.SE_BS(
S,
Simdur=P["dur"][x],
dt=P["dt"][x],
G_EoH=P["G_EoH"][x],
StochasticEoH=P["Stochastic_EoH"][x],
N=P["Nsyn"][x],
G_Dend=0.0,
L=P["L"][x],
D=P["D"][x],
constR=P["constR"][x],
gKLT_d=P["gKLT_d"][x],
gIH_d=P["gIH_d"][x],
gLEAK_d=P["gLEAK_d"][x],
somaticsynapsesonly=P["synapselocationtype"][x],
)
RD = sem.SE_BS(
S,
Simdur=P["dur"][x],
dt=P["dt"][x],
G_EoH=P["G_EoH"][x],
StochasticEoH=P["Stochastic_EoH"][x],
N=P["Nsyn"][x],
G_Dend=P["gsyn"][x],
L=P["L"][x],
D=P["D"][x],
constR=P["constR"][x],
gKLT_d=P["gKLT_d"][x],
gIH_d=P["gIH_d"][x],
gLEAK_d=P["gLEAK_d"][x],
somaticsynapsesonly=P["synapselocationtype"][x],
)
Ev0 = seh.SimpleDetectAP(R0["Vm"], thr=-100, dt=P["dt"][x], LM=-20, RM=10)
EvD = seh.SimpleDetectAP(RD["Vm"], thr=-100, dt=P["dt"][x], LM=-20, RM=10)
APr0 = len(Ev0["PeakT"]) / (P["dur"][x] / 1000.0)
APrD = len(EvD["PeakT"]) / (P["dur"][x] / 1000.0)
VS0, phi0, R0, phases0 = seh.vectorstrength(
Ev0["PeakT"], P["freq"][x], [0.0, P["dur"][x]]
)
VSD, phiD, RD, phasesD = seh.vectorstrength(
EvD["PeakT"], P["freq"][x], [0.0, P["dur"][x]]
)
print((str(x)))
return [VS0, phi0, R0, APr0, VSD, phiD, RD, APrD]
def runonce(x, P):
####hard coded settings, rarely change these:
APdetect_thr = -75
####GENERATE SPIKE INPUTS (the spiketimes are always the same to improve comparability)
thisdur = (P["dur"][x] - 5.0) / 1000.0
N = P["Nsyn"][x] + 1
anf_num = (N, 0, 0)
if P["AdvZseed"]:
thisZseed = P["Zseed"] + x
else:
thisZseed = P["Zseed"]
if P["AdvEoHseed"]:
thisEoHseed = P["EoHseed"] + x
else:
thisEoHseed = P["EoHseed"]
SO = seh.create_sound(fs=100e3,
freq=P["freq"][x],
duration=thisdur,
dbspl=P["dB"][x])
SP = seh.create_spk(SO,
fs=100e3,
N=N,
cf=P["cf"][x],
seed=thisZseed,
anf_num=anf_num) # generate enough spiketrains
S = np.array(SP["spikes"] * 1000.0)
R0 = sem.SE_3D(
S,
Simdur=P["dur"][x],
dt=P["dt"][x],
G_EoH=P["G_EoH"][x],
StochasticEoH=P["Stochastic_EoH"][x],
N=P["Nsyn"][x],
G_Dend=P["gsyn"][x] * 0.0, # set dendritic synapses to zero
tau_Dend=P["tau_dendsyn"][x],
gKLT_d=P["gKLT_d"][x],
gIH_d=P["gIH_d"][x],
gLEAK_d=P["gLEAK_d"][x],
cell=P["cell"][x],
EoHseed=thisEoHseed,
somaticsynapsesonly=P["synapselocationtype"][x],
)
#
RD = sem.SE_3D(
S,
Simdur=P["dur"][x],
dt=P["dt"][x],
G_EoH=P["G_EoH"][x],
StochasticEoH=P["Stochastic_EoH"][x],
N=P["Nsyn"][x],
G_Dend=P["gsyn"][x],
tau_Dend=P["tau_dendsyn"][x],
gKLT_d=P["gKLT_d"][x],
gIH_d=P["gIH_d"][x],
gLEAK_d=P["gLEAK_d"][x],
cell=P["cell"][x],
EoHseed=thisEoHseed,
somaticsynapsesonly=P["synapselocationtype"][x],
)
Ev0 = seh.SimpleDetectAP(R0["Vm"],
thr=APdetect_thr,
dt=P["dt"][x],
LM=-20,
RM=10)
EvD = seh.SimpleDetectAP(RD["Vm"],
thr=APdetect_thr,
dt=P["dt"][x],
LM=-20,
RM=10)
APr0 = len(Ev0["PeakT"]) / (P["dur"][x] / 1000.0)
APrD = len(EvD["PeakT"]) / (P["dur"][x] / 1000.0)
VS0, phi0, Ray0, phases0 = seh.vectorstrength(Ev0["PeakT"], P["freq"][x],
[0.0, P["dur"][x]])
VSD, phiD, RayD, phasesD = seh.vectorstrength(EvD["PeakT"], P["freq"][x],
[0.0, P["dur"][x]])
print((str(x)))
return [VS0, phi0, Ray0, APr0, VSD, phiD, RayD, APrD]
def mkexampletraces():
"""
As per reviewer request we generate example traces for two extreme conditions,
one at N=32 and g=0.06 (negative effect on rate) and one at N=32 and g=0.03
(positive effect on rate).
"""
dur = 5000.0
dur_s = dur / 1000.0
dt = 0.01
freq = 1389.0
cf = 1501.0
dB = 65.0
zseed = 45453
eohseed = 34823
stocheoh = True
geoh = 0.55
L = 50.0
D = 3.0
N = 32
anf_num = (N + 1, 0, 0)
kltd = 0.0085
hd = 0.001
ld = 0.001
SO = seh.create_sound(fs=100e3, freq=freq, duration=dur_s, dbspl=dB)
SP = seh.create_spk(SO,
fs=100e3,
N=N + 1,
cf=cf,
seed=zseed,
anf_num=anf_num)
S = np.array(SP["spikes"] * 1000.0)
print(".")
RX1 = sem.SE_BS(
S,
Simdur=dur,
dt=dt,
G_EoH=geoh,
StochasticEoH=stocheoh,
N=N,
G_Dend=0.016,
L=L,
D=D,
constR=False,
gKLT_d=kltd,
gIH_d=hd,
gLEAK_d=ld,
EoHseed=eohseed,
)
print(".")
RX2 = sem.SE_BS(
S,
Simdur=dur,
dt=dt,
G_EoH=geoh,
StochasticEoH=stocheoh,
N=N,
G_Dend=0.048,
L=L,
D=D,
constR=False,
gKLT_d=kltd,
gIH_d=hd,
gLEAK_d=ld,
EoHseed=eohseed,
)
Ev1 = seh.SimpleDetectAP(RX1["Vm"], thr=-100, dt=dt, LM=-20, RM=10)
Ev2 = seh.SimpleDetectAP(RX2["Vm"], thr=-100, dt=dt, LM=-20, RM=10)
APr1 = len(Ev1["PeakT"]) / dur_s
APr2 = len(Ev2["PeakT"]) / dur_s
VS1, phi1, Ray1, phases1 = seh.vectorstrength(Ev1["PeakT"], freq,
[0.0, dur])
VS2, phi2, Ray2, phases2 = seh.vectorstrength(Ev2["PeakT"], freq,
[0.0, dur])
tx = np.linspace(0, dur - dt, int(dur / dt))
fhand = plt.figure()
ax1 = plt.subplot(211)
ax1.set_title("Rate=" + str(APr1) + ", VS=" + str(VS1))
ax2 = plt.subplot(212)
ax2.set_title("Rate=" + str(APr2) + ", VS=" + str(VS2))
ax1.plot(tx, RX1["Vm"], "k")
ax1.set_xlim((800.0, 1000.0))
ax2.plot(tx, RX2["Vm"], color="r", linewidth=0.5)
ax2.set_xlim((800.0, 1000.0))
ax1.plot(Ev1["PeakT"],
Ev1["PeakV"],
color="g",
marker=".",
linestyle="None")
ax2.plot(Ev2["PeakT"],
Ev2["PeakV"],
color="r",
marker=".",
linestyle="None")
plt.tight_layout()
return fhand
def runonce(x, cf=450.0, freq=500.0, dbspl=65.0, dur=1000.0):
thisdur = (dur - 5.0) / 1000.0
N = 32
anf_num = (N + 1, 0, 0)
SO = seh.create_sound(
fs=100e3,
freq=freq,
duration=thisdur,
dbspl=dbspl,
)
SP = seh.create_spk(SO,
fs=100e3,
N=N + 1,
cf=cf,
seed=65564 + x,
anf_num=anf_num)
S = np.array(SP["spikes"] * 1000.0)
R0 = sem.SE_BS(
S,
Simdur=dur,
dt=0.01,
G_EoH=0.055,
StochasticEoH=True,
EoHseed=917634 + x,
N=32,
G_Dend=0.0,
)
print(".")
RD = sem.SE_BS(
S,
Simdur=dur,
dt=0.01,
G_EoH=0.055,
StochasticEoH=True,
EoHseed=917634 + x,
N=32,
G_Dend=0.016,
)
print(".")
R03 = sem.SE_3D(
S,
Simdur=dur,
dt=0.01,
G_EoH=0.055,
StochasticEoH=True,
EoHseed=917634 + x,
N=32,
G_Dend=0.0,
cell=1,
)
print(".")
RD3 = sem.SE_3D(
S,
Simdur=dur,
dt=0.01,
G_EoH=0.055,
StochasticEoH=True,
EoHseed=917634 + x,
N=32,
G_Dend=0.064,
cell=1,
)
print(".")
Ev0 = seh.SimpleDetectAP(R0["Vm"], thr=-100, dt=0.01, LM=-20, RM=10)
EvD = seh.SimpleDetectAP(RD["Vm"], thr=-100, dt=0.01, LM=-20, RM=10)
Ev03 = seh.SimpleDetectAP(R03["Vm"], thr=-100, dt=0.01, LM=-20, RM=10)
EvD3 = seh.SimpleDetectAP(RD3["Vm"], thr=-100, dt=0.01, LM=-20, RM=10)
VS0, phi0, R0, phases0 = seh.vectorstrength(Ev0["PeakT"], freq, [0.0, dur])
VSD, phiD, RD, phasesD = seh.vectorstrength(EvD["PeakT"], freq, [0.0, dur])
VS03, phi03, R03, phases03 = seh.vectorstrength(Ev03["PeakT"], freq,
[0.0, dur])
VSD3, phiD3, RD3, phasesD3 = seh.vectorstrength(EvD3["PeakT"], freq,
[0.0, dur])
nAP0 = len(Ev0["PeakT"])
nAPD = len(EvD["PeakT"])
nAP03 = len(Ev03["PeakT"])
nAPD3 = len(EvD3["PeakT"])
ISI0 = np.diff(Ev0["PeakT"])
ISID = np.diff(EvD["PeakT"])
ISI03 = np.diff(Ev03["PeakT"])
ISID3 = np.diff(EvD3["PeakT"])
cycdurms = 1000.0 / freq
ncycles = (thisdur * 1000.0) / cycdurms
bins = np.arange(0.0, 25.0, cycdurms)
hist0, bin_edges0 = np.histogram(ISI0, bins)
histD, bin_edgesD = np.histogram(ISID, bins)
hist03, bin_edges03 = np.histogram(ISI03, bins)
histD3, bin_edgesD3 = np.histogram(ISID3, bins)
EI0 = hist0[0] / ncycles
EID = histD[0] / ncycles
EI03 = hist03[0] / ncycles
EID3 = histD3[0] / ncycles
return (nAP0, nAPD, VS0, VSD, EI0, EID, nAP03, nAPD3, VS03, VSD3, EI03,
EID3)
def runmain():
rep = 10
simdur = 1000.0
dt = 0.01
N = 33
freq = 1251.0 # 1251.0
cf = 1001.0 # 1501.0
db = 70
SO = seh.create_sound(fs=100e3,
freq=freq,
duration=(simdur * 0.6) / 1000.0,
dbspl=db)
Silence = seh.create_sound(fs=100e3,
freq=freq,
duration=(simdur * 0.4) / 1000.0,
dbspl=-10)
R32 = []
R0 = []
Ev0 = []
Fail0 = []
Ev32 = []
Fail32 = []
APr0 = []
APr32 = []
allVS0 = []
allVS32 = []
allPhi0 = []
allPhi32 = []
#
SIN = []
DIN = []
SR0 = []
DR0 = []
SF0 = []
DF0 = []
SR32 = []
DR32 = []
SF32 = []
DF32 = []
for thisrep in range(rep):
SP = seh.create_spk(
np.concatenate((SO, Silence)),
fs=100e3,
N=N,
cf=cf,
seed=65562 + thisrep,
anf_num=(N, 0, 0),
)
S = np.array(SP["spikes"] * 1000.0)
R32.append(
sem.SE_BS(
S,
Simdur=simdur,
dt=dt,
G_EoH=0.055,
StochasticEoH=True,
N=N,
G_Dend=0.016,
EoHseed=8347236 + thisrep,
))
R0.append(
sem.SE_BS(
S,
Simdur=simdur,
dt=dt,
G_EoH=0.055,
StochasticEoH=True,
N=N,
G_Dend=0.0,
EoHseed=8347236 + thisrep,
))
Ev0.append(
seh.SimpleDetectAP(R0[-1]["Vm"], thr=-100, dt=dt, LM=-20, RM=10))
Fail0.append(seh.SimpleGetFails(S[-1], Ev0[-1], R0[-1], simdur, dt))
Ev32.append(
seh.SimpleDetectAP(R32[-1]["Vm"], thr=-100, dt=dt, LM=-20, RM=10))
Fail32.append(seh.SimpleGetFails(S[-1], Ev32[-1], R32[-1], simdur, dt))
APr0.append(len(Ev0[-1]["PeakT"]) / (simdur / 1000.0))
APr32.append(len(Ev32[-1]["PeakT"]) / (simdur / 1000.0))
#
tVS0, tphi0, tRay0, tphases0 = seh.vectorstrength(
Ev0[-1]["PeakT"], freq, [0.0, simdur * 0.6])
tVS32, tphi32, tRay32, tphases32 = seh.vectorstrength(
Ev32[-1]["PeakT"], freq, [0.0, simdur * 0.6])
allVS0.append(tVS0)
allVS32.append(tVS32)
allPhi0.append(tphi0)
allPhi32.append(tphi32)
#
# generate more results for writing...
stimlen = simdur * 0.6
stimlen_s = (simdur * 0.6) / 1000.0
spontlen_s = (simdur - stimlen) / 1000.0
myEv0 = np.array(Ev0[-1]["PeakT"])
myFail0 = np.array(Fail0[-1]["PeakT"])
myEv32 = np.array(Ev32[-1]["PeakT"])
myFail32 = np.array(Fail32[-1]["PeakT"])
SIN.append(np.sum(S[-1] > stimlen) / spontlen_s)
DIN.append(np.sum(S[-1] < stimlen) / stimlen_s)
SR0.append(np.sum(myEv0 > stimlen) / spontlen_s)
DR0.append(np.sum(myEv0 < stimlen) / stimlen_s)
SF0.append(np.sum(myFail0 > stimlen) / spontlen_s)
DF0.append(np.sum(myFail0 < stimlen) / stimlen_s)
SR32.append(np.sum(myEv32 > stimlen) / spontlen_s)
DR32.append(np.sum(myEv32 < stimlen) / stimlen_s)
SF32.append(np.sum(myFail32 > stimlen) / spontlen_s)
DF32.append(np.sum(myFail32 < stimlen) / stimlen_s)
SIN = np.array(SIN)
DIN = np.array(DIN)
SR0 = np.array(SR0)
DR0 = np.array(DR0)
SF0 = np.array(SF0)
DF0 = np.array(DF0)
SR32 = np.array(SR32)
DR32 = np.array(DR32)
SF32 = np.array(SF32)
DF32 = np.array(DF32)
# Pool spiketimes and construct smooth cyclehist
Spt0 = np.array(())
for s0 in Ev0:
Spt0 = np.append(Spt0, s0["PeakT"])
Spt32 = np.array(())
for s32 in Ev32:
Spt32 = np.append(Spt32, s32["PeakT"])
VS0, phi0, Ray0, phases0 = seh.vectorstrength(Spt0, freq,
[0.0, simdur * 0.6])
VS32, phi32, Ray32, phases32 = seh.vectorstrength(Spt32, freq,
[0.0, simdur * 0.6])
T = np.linspace(0, simdur - dt, int(round(simdur / dt)))
sT = np.linspace(0, (simdur * 0.8) - dt, int(round((simdur * 0.8) / dt)))
print("######Statistical comparison#########")
print("Total number of AP: " + str(len(Spt0)) + " vs. " + str(len(Spt32)))
outtext1 = str(np.mean(APr0)) + "+/-" + str(round(np.std(APr0), 2))
outtext2 = str(np.mean(APr32)) + "+/-" + str(round(np.std(APr32), 2))
print("Mean AP rate: " + outtext1 + " vs. " + outtext2)
Tval, p = stats.ttest_rel(APr0, APr32)
print("T=" + str(round(Tval, 3)) + ", p=" + str(p) + ", N=" + str(rep))
outtext4 = str(np.mean(allVS0)) + "+/-" + str(round(np.std(allVS0), 2))
outtext5 = str(np.mean(allVS32)) + "+/-" + str(round(np.std(allVS32), 2))
print("Mean VS: " + outtext4 + " vs. " + outtext5)
Tval, p = stats.ttest_rel(allVS0, allVS32)
print("T=" + str(round(Tval, 3)) + ", p=" + str(p) + ", N=" + str(rep))
outtext6 = str(np.mean(allPhi0)) + "+/-" + str(round(np.std(allPhi0), 2))
outtext7 = str(np.mean(allPhi32)) + "+/-" + str(round(np.std(allPhi32), 2))
print("Mean Phi: " + outtext6 + " vs. " + outtext7)
Tval, p = stats.ttest_rel(allPhi0, allPhi32)
print("T=" + str(round(Tval, 3)) + ", p=" + str(p) + ", N=" + str(rep))
#
print("Driven input rate: " + str(round(np.mean(DIN), 2)) + "+/-" +
str(round(np.std(DIN), 2)))
outtext1 = str(round(np.mean(DR0), 2)) + "+/-" + str(round(np.std(DR0), 2))
outtext2 = str(round(np.mean(DR32), 2)) + "+/-" + str(
round(np.std(DR32), 2))
print("Mean driven AP rate: " + outtext1 + " vs. " + outtext2)
Tval, p = stats.ttest_rel(DR0, DR32)
print("T=" + str(round(Tval, 3)) + ", p=" + str(p) + ", N=" + str(rep))
#
print("Spont input rate: " + str(round(np.mean(SIN), 2)) + "+/-" +
str(round(np.std(SIN), 2)))
outtext1 = str(round(np.mean(SR0), 2)) + "+/-" + str(round(np.std(SR0), 2))
outtext2 = str(round(np.mean(SR32), 2)) + "+/-" + str(
round(np.std(SR32), 2))
print("Mean spont AP rate: " + outtext1 + " vs. " + outtext2)
Tval, p = stats.ttest_rel(SR0, SR32)
print("T=" + str(round(Tval, 3)) + ", p=" + str(p) + ", N=" + str(rep))
#
outtext1 = str(round(np.mean(DF0), 2)) + "+/-" + str(round(np.std(DF0), 2))
outtext2 = str(round(np.mean(DF32), 2)) + "+/-" + str(
round(np.std(DF32), 2))
print("Mean driven Fail rate: " + outtext1 + " vs. " + outtext2)
Tval, p = stats.ttest_rel(DF0, DF32)
print("T=" + str(round(Tval, 3)) + ", p=" + str(p) + ", N=" + str(rep))
#
outtext1 = str(round(np.mean(SF0), 2)) + "+/-" + str(round(np.std(SF0), 2))
outtext2 = str(round(np.mean(SF32), 2)) + "+/-" + str(
round(np.std(SF32), 2))
print("Mean spont Fail rate: " + outtext1 + " vs. " + outtext2)
Tval, p = stats.ttest_rel(SF0, SF32)
print("T=" + str(round(Tval, 3)) + ", p=" + str(p) + ", N=" + str(rep))
print("Exact values for traces shown:")
print("Driven AP rate: " + str(DR0[0]) + " vs. " + str(DR32[0]))
print("Driven Fail rate: " + str(DF0[0]) + " vs. " + str(DF32[0]))
print("Spont AP rate: " + str(SR0[0]) + " vs. " + str(SR32[0]))
print("Spont Fail rate: " + str(SF0[0]) + " vs. " + str(SF32[0]))
# draw panels
fwidth = 8.5
fh = plt.figure(figsize=(fwidth / 2.54, fwidth / 2.54))
fh.subplots_adjust(left=0.15,
bottom=0.125,
right=0.95,
top=0.95,
wspace=0.55,
hspace=0.55)
sp1 = fh.add_subplot(2, 3, (1, 2))
sp1.plot(T, R0[0]["Vm"], "k-", linewidth=0.5)
sp1.plot(Ev0[0]["PeakT"], Ev0[0]["PeakV"], "g.", markersize=2)
sp1.plot(Fail0[0]["PeakT"], Fail0[0]["PeakV"], "r.", markersize=2)
sp1.plot(S[-1], np.zeros((len(S[-1]), )) + 20.0, "b.", markersize=2)
sp1.set_ylabel("Vm (mV)")
sp1.set_xlabel("Time (ms)")
sp1.spines["top"].set_visible(False)
sp1.spines["right"].set_visible(False)
sp1.plot(T, (np.concatenate((SO, Silence)) * 75) - 75.0,
color="g",
linewidth=0.5)
#
sp1b = fh.add_subplot(2, 3, 3)
H0 = np.histogram(phases0, np.linspace(0, 1, 33))
sp1b.plot(H0[1][1:], H0[0], "k")
sp1b.set_ylabel("AP/bin")
sp1b.set_xlabel("Phase (cycles)")
sp1b.set_xticks((0, 0.5, 1))
sp1b.set_yticks((0, 60, 120))
sp1b.plot((phi0, phi0), (0, sp1b.get_ylim()[1]), "r-")
annotext = r"VS=" + str(round(VS0, 2)) + "\n" + r"$\phi$=" + str(
round(phi0, 2))
sp1b.annotate(annotext, xy=(phi0, sp1b.get_ylim()[1]), fontsize=5)
sp1b.spines["top"].set_visible(False)
sp1b.spines["right"].set_visible(False)
#
sp2 = fh.add_subplot(2, 3, (4, 5))
sp2.plot(T, R32[0]["Vm"], "k-", linewidth=0.5)
sp2.plot(Ev32[0]["PeakT"], Ev32[0]["PeakV"], "g.", markersize=2)
sp2.plot(Fail32[0]["PeakT"], Fail32[0]["PeakV"], "r.", markersize=2)
sp2.plot(S[-1], np.zeros((len(S[-1]), )) + 20.0, "b.", markersize=2)
sp2.plot(T, (np.concatenate((SO, Silence)) * 75) - 75.0,
color="g",
linewidth=0.5)
sp2.set_ylabel("Vm (mV)")
sp2.set_xlabel("Time (ms)")
sp2.spines["top"].set_visible(False)
sp2.spines["right"].set_visible(False)
#
sp2b = fh.add_subplot(2, 3, 6)
H32 = np.histogram(phases32, np.linspace(0, 1, 33))
sp2b.plot(H32[1][1:], H32[0], "k")
sp2b.set_ylabel("AP/bin")
sp2b.set_xlabel("Phase (cycles)")
sp2b.set_xticks((0, 0.5, 1))
sp2b.set_yticks((0, 60, 120))
sp2b.plot((phi32, phi32), (0, sp2b.get_ylim()[1]), "r-")
annotext = r"VS=" + str(round(VS32, 2)) + "\n" + r"$\phi$=" + str(
round(phi32, 2))
sp2b.annotate(annotext, xy=(phi32, sp2b.get_ylim()[1]), fontsize=5)
sp2b.spines["top"].set_visible(False)
sp2b.spines["right"].set_visible(False)
#
pp = PdfPages("./figs/Figure6_raw.pdf")
pp.savefig(dpi=600)
pp.close()
def mkexampletraces(minL, minD, maxL, maxD):
"""
As per reviewer request we generate example traces for extreme conditions
"""
dur = 5000.0
dur_s = dur / 1000.0
dt = 0.01
freq = 1389.0
cf = 1501.0
dB = 65.0
zseed = 45453
eohseed = 34823
stocheoh = True
geoh = 0.055
N = 32
gsyn = 0.016
anf_num = (N + 1, 0, 0)
SO = seh.create_sound(fs=100e3, freq=freq, duration=dur_s, dbspl=dB)
SP = seh.create_spk(SO,
fs=100e3,
N=N + 1,
cf=cf,
seed=zseed,
anf_num=anf_num)
S = np.array(SP["spikes"] * 1000.0)
print(".")
RX1 = sem.SE_BS(
S,
Simdur=dur,
dt=dt,
G_EoH=geoh,
StochasticEoH=stocheoh,
N=N,
G_Dend=0.0,
L=minL,
D=minD,
constR=False,
EoHseed=eohseed,
)
print(".")
RX2 = sem.SE_BS(
S,
Simdur=dur,
dt=dt,
G_EoH=geoh,
StochasticEoH=stocheoh,
N=N,
G_Dend=gsyn,
L=minL,
D=minD,
constR=False,
EoHseed=eohseed,
)
print(".")
RX3 = sem.SE_BS(
S,
Simdur=dur,
dt=dt,
G_EoH=geoh,
StochasticEoH=stocheoh,
N=N,
G_Dend=0.0,
L=maxL,
D=maxD,
constR=False,
EoHseed=eohseed,
)
print(".")
RX4 = sem.SE_BS(
S,
Simdur=dur,
dt=dt,
G_EoH=geoh,
StochasticEoH=stocheoh,
N=N,
G_Dend=gsyn,
L=maxL,
D=maxD,
constR=False,
EoHseed=eohseed,
)
Ev1 = seh.SimpleDetectAP(RX1["Vm"], thr=-100, dt=dt, LM=-20, RM=10)
Ev2 = seh.SimpleDetectAP(RX2["Vm"], thr=-100, dt=dt, LM=-20, RM=10)
Ev3 = seh.SimpleDetectAP(RX3["Vm"], thr=-100, dt=dt, LM=-20, RM=10)
Ev4 = seh.SimpleDetectAP(RX4["Vm"], thr=-100, dt=dt, LM=-20, RM=10)
APr1 = np.round(len(Ev1["PeakT"]) / dur_s)
APr2 = np.round(len(Ev2["PeakT"]) / dur_s)
APr3 = np.round(len(Ev3["PeakT"]) / dur_s)
APr4 = np.round(len(Ev4["PeakT"]) / dur_s)
VS1, phi1, Ray1, phases1 = seh.vectorstrength(Ev1["PeakT"], freq,
[0.0, dur])
VS2, phi2, Ray2, phases2 = seh.vectorstrength(Ev2["PeakT"], freq,
[0.0, dur])
VS3, phi3, Ray3, phases3 = seh.vectorstrength(Ev3["PeakT"], freq,
[0.0, dur])
VS4, phi4, Ray4, phases4 = seh.vectorstrength(Ev4["PeakT"], freq,
[0.0, dur])
tx = np.linspace(0, dur - dt, int(dur / dt))
fhand = plt.figure()
ax1 = plt.subplot(211)
ax1.set_title(
str(APr1) + ", " + str(VS1) + " / " + str(APr2) + ", " + str(VS2))
ax1.plot(tx, RX1["Vm"], "k")
ax1.plot(Ev1["PeakT"],
Ev1["PeakV"],
color="g",
marker=".",
linestyle="None")
ax1.plot(tx, RX2["Vm"], "r")
ax1.plot(Ev2["PeakT"],
Ev2["PeakV"],
color="r",
marker=".",
linestyle="None")
ax1.set_xlim((300.0, 500.0))
ax2 = plt.subplot(212)
ax2.set_title(
str(APr3) + ", " + str(VS3) + " / " + str(APr4) + ", " + str(VS4))
ax2.plot(tx, RX3["Vm"], color="k", linewidth=0.5)
ax2.plot(Ev3["PeakT"],
Ev3["PeakV"],
color="g",
marker=".",
linestyle="None")
ax2.plot(tx, RX4["Vm"], color="r", linewidth=0.5)
ax2.plot(Ev4["PeakT"],
Ev4["PeakV"],
color="r",
marker=".",
linestyle="None")
ax2.set_xlim((300.0, 500.0))
plt.tight_layout()
return fhand