Ejemplo n.º 1
0
Archivo: cK.py Proyecto: g13/bNEURON
def cK(cid, seed, fmt, run_t, tstep, rE, rI, testLinear, testBilinear,
       plotSindivid, plotBindivid, newSv, sender):
    directory = str(seed)
    g0 = 32.0 * 6e-4
    #testLinear = True

    # initialize cell
    #locE = np.array([60],dtype='int')
    #locI = np.array([14],dtype='int')
    locE = np.array([60, 72, 78, 84, 90, 98], dtype='int')
    locI = np.array([14, 28, 30], dtype='int')
    #gE = np.array([1.0, 1.0, 1.0, 1.0, 1.0, 1.0]) * g0
    #gI = -g0*np.array([10.0, 10.0, 10.0])

    #locE = np.array([79, 82, 83, 98, 120, 124],dtype='int')
    #locI = np.array([14, 28, 40],dtype='int')
    gE = np.array([0.6, 0.6, 0.2, 0.6, 0.15, 0.6]) * g0
    gI = -g0 * np.array([6.0, 10.0, 8.0])

    posE = np.array([0.3, 0.3, 0.9, 0.6, 0.4, 0.2])
    posI = np.array([0.7, 0.2, 0.5])
    gE = gE[:locE.size]
    gI = gI[:locI.size]
    posE = posE[:locE.size]
    posI = posI[:locI.size]
    pos = np.concatenate((posE, posI))
    loc = np.concatenate((locE, locI))
    gList = np.concatenate((gE, gI))
    vrest = -70.0
    vThres = -54.0
    alphaR = True

    run_nt = int(round(run_t / tstep)) + 1
    n = loc.size

    cell, vSL, sL, _ = prepCell(gList, loc, pos, n, vrest, alphaR)
    # population events

    rE = rE / 1000.0
    rI = rI / 1000.0
    rates = np.empty(n)
    rates[:locE.size] = np.random.randn(locE.size) * rE * 0.1 + rE
    rates[locE.size:] = np.random.randn(locI.size) * rI * 0.1 + rI
    print rates
    np.random.seed(seed)
    spikeTrain = np.empty(n, dtype=object)
    IDs = np.empty(n, dtype=object)
    activePeriod = run_t / 3.0
    for i in xrange(n):
        if rates[i] <= 0.0:
            spikeTrain[i] = np.array([run_t + 1])
            continue
        ts = 0
        events = []
        while ts < activePeriod:
            tt = np.random.poisson(1.0 / rates[i], 1)
            if tt > 0:
                ts += int(round(tt / tstep)) * tstep
            events.append(ts)
        print str(i) + 'th synapse:', len(events), 'spikes'
        spikeTrain[i] = np.array(events)
    RList = np.zeros((n, 2))
    dendVclamp = np.zeros(n) + 1000
    for i in xrange(n):
        IDs[i] = i + np.zeros(spikeTrain[i].size, dtype=int)
    IDs = np.hstack(IDs)
    spikes = np.hstack(spikeTrain)
    indSorted = np.argsort(spikes)
    sortedID = IDs[indSorted]
    sortedSpikes = spikes[indSorted]
    itsp = np.round(sortedSpikes / tstep).astype(int)

    if testLinear:
        plotLinear = True
    else:
        plotLinear = False
    if testBilinear:
        plotBilinear = True
    else:
        plotBilinear = False
    if not os.path.exists(directory):
        os.mkdir(directory)
    directory = directory + '/'
    # run simulation
    v0 = vrest
    trans = 0
    t0 = 0
    oneGo = True
    getDendV = True
    monitorDend = False
    pas = False
    plotSlice = True
    tref = 10.0
    vBack = vThres - 2.0
    simv, _, _, _, dendv = proceed(cell,
                                   v0,
                                   sL,
                                   RList,
                                   vSL,
                                   spikeTrain,
                                   n,
                                   trans,
                                   run_t,
                                   vBack,
                                   tref,
                                   vThres,
                                   oneGo,
                                   t0,
                                   tstep,
                                   loc,
                                   pos,
                                   dendVclamp,
                                   alphaR,
                                   getDendV,
                                   monitorDend,
                                   pas,
                                   plotSlice,
                                   fign=directory + 'simvOnly')

    # run linear
    if testLinear:
        liv = np.zeros(run_nt, dtype=float) + vrest
        liv0 = np.zeros(run_nt, dtype=float) + vrest
        liv1 = np.zeros(run_nt, dtype=float) + vrest
        tol = True
        basic_run_t = 1
        for i in xrange(sortedSpikes.size):
            it = itsp[i]
            sel = [sortedID[i]]
            if sortedSpikes[i] >= run_t:
                break
            if i == 0:
                v0 = vrest
            else:
                v0 = liv[it]
            ## liv
            tsp = np.array([[0]])
            cpi = False
            sv, _, dendv = bproceed(cell, v0, vThres, sL, gList, vSL, tsp, n,
                                    sel, basic_run_t, tstep, '', pos, loc,
                                    alphaR, cpi, tol)
            if it + sv.size >= run_nt:
                et = run_nt - it
            else:
                et = sv.size
            sv = sv[:et] - v0
            dendv = dendv[:, :et] - v0
            liv[it:it + et] += sv
            if plotSindivid:
                fign = directory + 'liv' + str(i)
                fig = pyplot.figure(fign, figsize=(8, 4))
                ax = fig.add_subplot(1, 1, 1)
                ax.plot(np.arange(sv.size) * tstep, sv)
                ax.plot(np.arange(sv.size) * tstep, dendv.T)
                pyplot.savefig(fign + '.' + fmt, format=fmt, dpi=900)
            ## liv0
            tsp = np.array([[0]])
            cpi = False
            v0 = liv0[it]
            sv, _, dendv = bproceed(cell, vrest, vThres, sL, gList, vSL, tsp,
                                    n, sel, basic_run_t, tstep, '', pos, loc,
                                    alphaR, cpi, tol)
            if it + sv.size >= run_nt:
                et = run_nt - it
            else:
                et = sv.size
            sv = sv[:et] - vrest
            dendv = dendv[:, :et] - vrest
            liv0[it:it + et] += sv
            if plotSindivid:
                fign = directory + 'liv0' + str(i)
                fig = pyplot.figure(fign, figsize=(8, 4))
                ax = fig.add_subplot(1, 1, 1)
                ax.plot(np.arange(sv.size) * tstep, sv)
                ax.plot(np.arange(sv.size) * tstep, dendv.T)
                pyplot.savefig(fign + '.' + fmt, format=fmt, dpi=900)
            ## liv1
            tsp = np.array([[0]])
            v0 = liv1[it]
            print 'liv1' + '-' + str(i), v0
            cpi = True
            cell.set_volt(v0)
            sv, _, dendv = bproceed(cell, vrest, vThres, sL, gList, vSL, tsp,
                                    n, sel, basic_run_t, tstep, '', pos, loc,
                                    alphaR, cpi, tol)
            if it + sv.size >= run_nt:
                et = run_nt - it
            else:
                et = sv.size
            sel = []
            tsp = np.array([[]])
            leaky_run_t = (et - 1) * tstep
            leakyV, _, leakyDendV = bproceed(cell, vrest, vThres, sL, gList,
                                             vSL, tsp, n, sel, leaky_run_t,
                                             tstep, '', pos, loc, alphaR, cpi)
            sv = sv[:et] - leakyV
            dendv = dendv[:, :et] - leakyDendV
            liv1[it:it + et] += sv
            if plotSindivid:
                fign = directory + 'liv1' + str(i)
                fig = pyplot.figure(fign, figsize=(8, 4))
                ax = fig.add_subplot(1, 1, 1)
                ax.plot(np.arange(sv.size) * tstep, sv)
                ax.plot(np.arange(sv.size) * tstep, dendv.T)
                pyplot.savefig(fign + '.' + fmt, format=fmt, dpi=900)

# run bilinear
    if testBilinear:
        tol = True
        biv = np.zeros(run_nt, dtype=float) + vrest
        biv0 = np.zeros(run_nt, dtype=float) + vrest
        biv1 = np.zeros(run_nt, dtype=float) + vrest
        nonv = np.zeros(sortedSpikes.size, dtype=float) + vrest
        nonv1 = np.zeros(sortedSpikes.size, dtype=float) + vrest
        sv = np.empty(sortedSpikes.size, dtype=object)
        sv0 = np.empty(sortedSpikes.size, dtype=object)
        sv1 = np.empty(sortedSpikes.size, dtype=object)
        for i in xrange(sortedSpikes.size):
            if sortedSpikes[i] >= run_t:
                break
            it = itsp[i]
            sel = [sortedID[i]]
            if i == 0:
                v0 = vrest
            else:
                v0 = biv[it]
            nonv[i] = v0
            cpi = False
            tsp = np.array([[0]])
            sv[i], _, dendv = bproceed(cell, v0, vThres, sL, gList, vSL, tsp,
                                       n, sel, run_t, tstep, '', pos, loc,
                                       alphaR, cpi, tol)
            if it + sv[i].size >= run_nt:
                et = run_nt - it
            else:
                et = sv[i].size
            sv[i] = sv[i][:et] - v0
            dendv = dendv[:, :et] - v0
            biv[it:it + et] += sv[i]
            if plotSindivid:
                fign = directory + 'biv_' + str(i)
                fig = pyplot.figure(fign, figsize=(8, 4))
                ax = fig.add_subplot(1, 1, 1)
                ax.plot(np.arange(sv[i].size) * tstep, sv[i])
                ax.plot(np.arange(sv[i].size) * tstep, dendv.T)
                pyplot.savefig(fign + '.' + fmt, format=fmt, dpi=900)

            for j in xrange(i):
                jt = itsp[j]
                if jt + sv[j].size < it:
                    continue
                if sortedID[j] == sortedID[i]:
                    sel = [sortedID[j]]
                    tsp = np.array([[0, sortedSpikes[i] - sortedSpikes[j]]])
                else:
                    sel = [sortedID[j], sortedID[i]]
                    tsp = np.array([[0], [sortedSpikes[i] - sortedSpikes[j]]])
                idt = it - jt
                b_run_t = sv[j].size * tstep
                v0 = nonv[j]
                bv, _, dendv = bproceed(cell, v0, vThres, sL, gList, vSL, tsp,
                                        n, sel, b_run_t, tstep, '', pos, loc,
                                        alphaR, cpi, tol)
                if jt + bv.size >= run_nt:
                    et = run_nt - jt
                else:
                    et = bv.size
                bv = bv[:et] - v0
                dendv = dendv[:, :et] - v0
                if sv[j].size > et:
                    s1t = et
                else:
                    s1t = sv[j].size

                if newSv:
                    sel = [sortedID[i]]
                    tsp = np.array([[0]])
                    sv2, _, dendv = bproceed(cell, v0, vThres, sL, gList, vSL,
                                             tsp, n, sel, (et - idt) * tstep,
                                             tstep, '', pos, loc, alphaR, cpi)
                    sv2 = sv2 - v0
                else:
                    sv2 = sv[i]

                if et - idt > sv2.size:
                    s2t = sv2.size
                else:
                    s2t = et - idt
                print 's2:', it, ', ', s2t, '<=', sv2.size
                print 's1:', jt, ', ', s1t, '<=', sv[j].size, 'kv start:', idt
                kv = bv.copy()
                kv[:s1t] -= sv[j][:s1t]
                kv[idt:idt + s2t] -= sv2[:s2t]
                addv = np.zeros(bv.size)
                addv[:s1t] += sv[j][:s1t]
                addv[idt:idt + s2t] += sv2[:s2t]

                biv[jt:jt + et] += kv

                if plotBindivid:
                    fign = directory + 'biv' + str(i) + '-' + str(j)
                    fig = pyplot.figure(fign, figsize=(8, 4))
                    ax = fig.add_subplot(1, 1, 1)
                    t = np.arange(bv.size) * tstep
                    ax.plot(t, bv, 'b')
                    #ax.plot(t,dendv.T)
                    ax.plot(t, kv, ':k')
                    ax.plot(t, addv, ':g')
                    ax.plot(t[:s1t], sv[j][:s1t], ':r')
                    ax.plot(t[idt:idt + s2t], sv2[:s2t], ':b')
                    pyplot.savefig(fign + '.' + fmt, format=fmt, dpi=900)

            v0 = biv0[it]
            sel = [sortedID[i]]
            tsp = np.array([[0]])
            sv0[i], _, dendv = bproceed(cell, vrest, vThres, sL, gList, vSL,
                                        tsp, n, sel, run_t, tstep, '', pos,
                                        loc, alphaR, cpi, tol)
            if it + sv0[i].size >= run_nt:
                et = run_nt - it
            else:
                et = sv0[i].size
            sv0[i] = sv0[i][:et] - vrest
            dendv = dendv[:, :et] - vrest
            biv0[it:it + et] += sv0[i]
            if plotSindivid:
                fign = directory + 'biv0_' + str(i)
                fig = pyplot.figure(fign, figsize=(8, 4))
                ax = fig.add_subplot(1, 1, 1)
                ax.plot(np.arange(sv0[i].size) * tstep, sv0[i])
                ax.plot(np.arange(sv0[i].size) * tstep, dendv.T)
                pyplot.savefig(fign + '.' + fmt, format=fmt, dpi=900)

            for j in xrange(i):
                jt = itsp[j]
                if jt + sv0[j].size < it:
                    continue
                if sortedID[j] == sortedID[i]:
                    sel = [sortedID[j]]
                    tsp = np.array([[0, sortedSpikes[i] - sortedSpikes[j]]])
                else:
                    sel = [sortedID[j], sortedID[i]]
                    tsp = np.array([[0], [sortedSpikes[i] - sortedSpikes[j]]])
                idt = it - jt
                b_run_t = sv0[j].size * tstep
                bv, _, dendv = bproceed(cell, vrest, vThres, sL, gList, vSL,
                                        tsp, n, sel, b_run_t, tstep, '', pos,
                                        loc, alphaR, cpi, tol)
                if jt + bv.size >= run_nt:
                    et = run_nt - jt
                else:
                    et = bv.size
                bv = bv[:et] - vrest
                dendv = dendv[:, :et] - vrest
                if sv0[j].size > et:
                    s1t = et
                else:
                    s1t = sv0[j].size

                sv2 = sv0[i]

                if et - idt > sv2.size:
                    s2t = sv2.size
                else:
                    s2t = et - idt
                print 's2:', it, ', ', s2t, '<=', sv2.size
                print 's1:', jt, ', ', s1t, '<=', sv0[j].size, 'kv start:', idt
                kv = bv.copy()
                kv[:s1t] -= sv0[j][:s1t]
                kv[idt:idt + s2t] -= sv2[:s2t]
                addv = np.zeros(bv.size)
                addv[:s1t] += sv0[j][:s1t]
                addv[idt:idt + s2t] += sv2[:s2t]

                biv0[jt:jt + et] += kv

                if plotBindivid:
                    fign = directory + 'biv0_' + str(i) + '-' + str(j)
                    fig = pyplot.figure(fign, figsize=(8, 4))
                    ax = fig.add_subplot(1, 1, 1)
                    t = np.arange(bv.size) * tstep
                    ax.plot(t, bv, 'b')
                    #ax.plot(t,dendv.T)
                    ax.plot(t, kv, ':k')
                    ax.plot(t, addv, ':g')
                    ax.plot(t[:s1t], sv0[j][:s1t], ':r')
                    ax.plot(t[idt:idt + s2t], sv2[:s2t], ':b')
                    pyplot.savefig(fign + '.' + fmt, format=fmt, dpi=900)

            sel = [sortedID[i]]
            if i == 0:
                v0 = vrest
            else:
                v0 = biv1[it]
            nonv1[i] = v0
            cpi = True
            cell.set_volt(v0)
            tsp = np.array([[0]])
            sv1[i], _, dendv = bproceed(cell, v0, vThres, sL, gList, vSL, tsp,
                                        n, sel, run_t, tstep, '', pos, loc,
                                        alphaR, cpi, tol)
            if it + sv1[i].size >= run_nt:
                et = run_nt - it
            else:
                et = sv1[i].size
            leaky_run_t = (et - 1) * tstep
            sel = []
            tsp = np.array([[]])
            cell.set_volt(v0)
            leakyV, _, leakyDendV = bproceed(cell, v0, vThres, sL, gList, vSL,
                                             tsp, n, sel, leaky_run_t, tstep,
                                             '', pos, loc, alphaR, cpi)
            sv1[i] = sv1[i][:et] - leakyV
            dendv = dendv[:, :et] - leakyDendV
            biv1[it:it + et] += sv1[i]
            if plotSindivid:
                fign = directory + 'biv1_' + str(i)
                fig = pyplot.figure(fign, figsize=(8, 4))
                ax = fig.add_subplot(1, 1, 1)
                ax.plot(np.arange(sv1[i].size) * tstep, sv1[i])
                ax.plot(np.arange(sv1[i].size) * tstep, dendv.T)
                pyplot.savefig(fign + '.' + fmt, format=fmt, dpi=900)

            for j in xrange(i):
                jt = itsp[j]
                if jt + sv1[j].size < it:
                    continue
                if sortedID[j] == sortedID[i]:
                    sel = [sortedID[j]]
                    tsp = np.array([[0, sortedSpikes[i] - sortedSpikes[j]]])
                else:
                    sel = [sortedID[j], sortedID[i]]
                    tsp = np.array([[0], [sortedSpikes[i] - sortedSpikes[j]]])
                idt = it - jt
                b_run_t = sv1[j].size * tstep
                v0 = nonv1[j]
                cell.set_volt(v0)
                bv, _, dendv = bproceed(cell, v0, vThres, sL, gList, vSL, tsp,
                                        n, sel, b_run_t, tstep, '', pos, loc,
                                        alphaR, cpi, tol)
                if jt + bv.size >= run_nt:
                    et = run_nt - jt
                else:
                    et = bv.size
                leaky_run_t = (et - 1) * tstep
                sel = []
                tsp = np.array([[]])
                cell.set_volt(v0)
                leakyV, _, leakyDendV = bproceed(cell, v0, vThres, sL, gList,
                                                 vSL, tsp, n, sel, leaky_run_t,
                                                 tstep, '', pos, loc, alphaR,
                                                 cpi)
                bv = bv[:et] - leakyV
                dendv = dendv[:, :et] - leakyDendV
                if sv1[j].size > et:
                    s1t = et
                else:
                    s1t = sv1[j].size

                if newSv:
                    sel = [sortedID[i]]
                    tsp = np.array([[0]])
                    v0 = bv[idt]
                    cell.set_volt(v0)
                    sv2, _, dendv = bproceed(cell, v0, vThres, sL, gList, vSL,
                                             tsp, n, sel, (et - idt) * tstep,
                                             tstep, '', pos, loc, alphaR, cpi)
                    leaky_run_t = (sv2.size - 1) * tstep
                    sel = []
                    tsp = np.array([[]])
                    cell.set_volt(v0)
                    leakyV, _, leakyDendV = bproceed(cell, v0, vThres, sL,
                                                     gList, vSL, tsp, n, sel,
                                                     leaky_run_t, tstep, '',
                                                     pos, loc, alphaR, cpi)
                    sv2 = sv2 - leakyV
                else:
                    sv2 = sv1[i]

                if et - idt > sv2.size:
                    s2t = sv2.size
                else:
                    s2t = et - idt
                print 's2:', it, ', ', s2t, '<=', sv2.size
                print 's1:', jt, ', ', s1t, '<=', sv1[j].size, 'kv start:', idt
                kv = bv.copy()
                kv[:s1t] -= sv1[j][:s1t]
                kv[idt:idt + s2t] -= sv2[:s2t]
                addv = np.zeros(bv.size)
                addv[:s1t] += sv1[j][:s1t]
                addv[idt:idt + s2t] += sv2[:s2t]

                biv1[jt:jt + et] += kv

                if plotBindivid:
                    fign = directory + 'biv1_' + str(i) + '-' + str(j)
                    fig = pyplot.figure(fign, figsize=(8, 4))
                    ax = fig.add_subplot(1, 1, 1)
                    t = np.arange(bv.size) * tstep
                    ax.plot(t, bv, 'b')
                    #ax.plot(t,dendv.T)
                    ax.plot(t, kv, ':k')
                    ax.plot(t, addv, ':g')
                    ax.plot(t[:s1t], sv1[j][:s1t], ':r')
                    ax.plot(t[idt:idt + s2t], sv2[:s2t], ':b')
                    pyplot.savefig(fign + '.' + fmt, format=fmt, dpi=900)


# plot and compare
    t = np.arange(run_nt) * tstep
    fign = 'compare' + str(seed)
    fig = pyplot.figure(fign, figsize=(8, 4))
    ax1 = fig.add_subplot(2, 1, 1)
    ax1.plot(t, simv, 'k')
    if plotLinear:
        ax1.plot(t, liv, 'r')
        ax1.plot(t, liv0, 'g')
        ax1.plot(t, liv1, 'm')
    if plotBilinear:
        ax1.plot(t, biv, 'b')
        ax1.plot(t, biv0, 'c')
        ax1.plot(t, biv1, 'y')

    ax2 = fig.add_subplot(2, 1, 2)
    if plotLinear:
        ax2.plot(t, simv - liv, ':r')
        ax2.plot(t, simv - liv0, ':g')
        ax2.plot(t, simv - liv1, ':m')
    if plotBilinear:
        ax2.plot(t, simv - biv, ':b')
        ax2.plot(t, simv - biv0, ':c')
        ax2.plot(t, simv - biv1, ':y')

    pyplot.savefig(fign + '.' + fmt, format=fmt, dpi=900)
    datafn = directory + 'data.bin'
    write_one(datafn, np.array([np.int(run_nt)]), 'wb')
    write_one(datafn, simv)
    write_one(datafn, np.array([np.int(testLinear)]))
    if testLinear:
        write_one(datafn, liv)
        write_one(datafn, liv0)
        write_one(datafn, liv1)
    write_one(datafn, np.array([np.int(testBilinear)]))
    if testBilinear:
        write_one(datafn, biv)
        write_one(datafn, biv0)
        write_one(datafn, biv1)
    print cid, 'finished'
    sender.send(np.array([cid, datafn], dtype=object))
Ejemplo n.º 2
0
    #gI = -g0*np.array([6.0, 10.0, 8.0])

    posE = np.array([0.3,0.3,0.9,0.6,0.4,0.2])
    posI = np.array([0.7,0.2,0.5])
    gE = gE[:locE.size]
    gI = gI[:locI.size]
    posE = posE[:locE.size]
    posI = posI[:locI.size]
    pos = np.concatenate((posE, posI))
    loc = np.concatenate((locE, locI))
    gList = np.concatenate((gE, gI))
    vrest = -70.0
    vThres = -60.0
    n = loc.size
    alphaR = True 
    cell, vSL, synList, _ = prepCell(gList, loc, pos, n, vrest, alphaR)
    
    #vRange = np.array([-74,-70,-66,-62],dtype='double')
    vRange = np.array([-74,-70,-66],dtype='double')
    dt = 10.0
    theme = 'test'
    datafn = 'leakySingle-' + theme + '.npy'
    # linear and leaky
    plotLeakySingle = False 
    t = np.arange(run_nt)*tstep
    cpi = True
    nv = vRange.size
    if not loadSingle:
        V = np.empty((run_nt,n,nv))
        dendV = np.empty((run_nt,n,n,nv))
        tMax = np.empty((n,nv))
Ejemplo n.º 3
0
    seed = 231271 
    np.random.seed(seed)
    locE = np.array([60, 72, 78, 84, 90, 98],dtype='int')
    locI = np.array([14, 28, 30],dtype='int')
    g0 = 32.0*5e-4
    gE = np.array([1.0, 1.0, 1.0, 1.0, 1.0, 1.0]) * g0
    gI = -g0*np.array([10.0, 10.0, 10.0])
    posE = np.array([0.3,0.3,0.9,0.6,0.4,0.2])
    posI = np.array([0.7,0.2,0.5])
    pos = np.concatenate((posE, posI))
    loc = np.concatenate((locE, locI))
    gList = np.concatenate((gE, gI))
    v0 = -70
    n = loc.size
    alphaR = True 
    cell, vecStimList, synList, distance = prepCell(gList, loc, pos, n, v0, alphaR)

    sel = range(n)

    vecTuple = [np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1])]
    #vecTuple = (np.array([0,330,run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([run_t+1]),np.array([0,660,run_t+1]))
    #RList = np.array([[0, 0], [0, 0], [0, 0], [0, 0], [0, 0], [0, 0], [0, 0], [0, 0], [0, 0]],dtype='double')
    RList0 = np.zeros((9,2))
    #dendVclamp = np.array([1000, 1000, 1000, 1000, 1000, 1000, 1000, 1000, 1000])
    dendVclamp = np.array([-50, -50, -50, -50, -50, -50, 1000, 1000, 1000])
    t0 = 0.0
    tref = 13
    vThres = -65.0
    #v_pre = -70
    v_pre = vThres
    vBack = -61.0
Ejemplo n.º 4
0
Archivo: sv.py Proyecto: g13/bNEURON
def generateSvData(datafn,alphaR,vRange,gList,loc,pos,vrest,vThres,run_t,tstep,fmt,theme,normVrest=False):
    run_nt = int(round(run_t/tstep))+1
    nv = vRange.size
    n = loc.size
    cell, vSL, sL, _ = prepCell(gList, loc, pos, n, vrest, alphaR)
    V = np.empty((nv,n,run_nt))
    dendV = np.empty((nv,n,n,run_nt))
    tMax = np.empty((nv,n),dtype='int')
    sFire = np.empty((nv,n),dtype='int')
    if normVrest:
        leakyV = np.empty((nv,run_nt))
        leakyDendV = np.empty((nv,n,run_nt))
        print 'get Leaky'
        jobs = []
        receivers = []
        for vi in xrange(nv):
            receiver, sender = mp.Pipe(False)
            job = mp.Process(target=getLeaky, args = (vi,vRange,cell,vThres,gList,loc,pos,sL,vSL,n,tstep,run_t,alphaR,sender))
            jobs.append(job)
            receivers.append(receiver)
            job.start()
        gather_and_distribute_results(receivers, jobs, getLeaky.__name__, nv, 0, leakyV, leakyDendV)
        fign = 'leakyV' + theme 
        fig = pyplot.figure(fign, figsize=(8,4))
        ax = fig.add_subplot(1,1,1)
        ax.plot(np.arange(run_nt)*tstep,leakyV.T,'k')
        prop_cycle = pyplot.rcParams['axes.prop_cycle']
        colors = prop_cycle.by_key()['color']
        for vi in xrange(nv):
            ax.set_prop_cycle('color',colors)
            ax.plot(np.arange(run_nt)*tstep,leakyDendV[vi,:,:].T,':')
        pyplot.savefig(fign+'.'+fmt,format=fmt,bbox_inches='tight',dpi=900)

    print 'get singlets'
    jobs = []
    receivers = []
    for vi in xrange(nv):
        receiver, sender = mp.Pipe(False)
        if normVrest:
            lV = leakyV[vi,:]
            lDV = leakyDendV[vi,:]
        else:
            lV = []
            lDV = []
        job = mp.Process(target=getSingletsV, args = (vi,vRange,cell,vThres,gList,loc,pos,sL,vSL,n,tstep,run_t,run_nt,alphaR,sender,normVrest,lV,lDV,fmt))
        jobs.append(job)
        receivers.append(receiver)
        job.start()
    print 'waiting for all jobs'
    try:
        gather_and_distribute_results(receivers, jobs, getSingletsV.__name__, nv, 0, V, dendV, tMax, sFire)
    except EOFError:
        print "尼玛"
    else:
        print "finished"

    if normVrest:
        write_one(datafn+'.bin',V,mode='wb')
        write_one(datafn+'.bin',tMax,mode='ab')
        write_one(datafn+'.bin',sFire,mode='ab')
        write_one(datafn+'.bin',leakyV,mode='ab')
        write_one(datafn+'.bin',dendV,mode='ab')
        write_one(datafn+'.bin',leakyDendV,mode='ab')
        #with open(datafn+'.npz', 'w') as leakySingleData:
        #    np.savez(leakySingleData, V0=V, dendV0=dendV, tMax0=tMax, sFire0=sFire, leakyV0=leakyV, leakyDendV0=leakyDendV)
    else:
        write_one(datafn+'.bin',V,mode='wb')
        write_one(datafn+'.bin',tMax,mode='ab')
        write_one(datafn+'.bin',sFire,mode='ab')
        write_one(datafn+'.bin',dendV,mode='ab')
        #with open(datafn+'.npz', 'w') as leakySingleData:
        #    np.savez(leakySingleData, V=V, dendV=dendV, tMax=tMax, sFire=sFire)
    return V, tMax