def show_rates(d, **k):
scale=k.get('scale',2)
kw={'n_rows':1,
'n_cols':5,
'w':int(72/2.54*17.6)*2*scale,
'h':int(72/2.54*17.6)*1.5*scale,
'fontsize':7*scale,
'title_fontsize':7*scale,
'frame_hight_y':0.5,
'frame_hight_x':0.7,
'gs_builder':gs_builder}
fig, axs=ps.get_figure2(**kw)
i=0
images=[]
for tp in ['beta', 'sw']:
for net in sorted(d[tp].keys()):
ax=axs[i]
dd=d[tp][net]
im=plot_rates(ax, dd['mean_rates'], dd['xlabels'], dd['ylabels'])
ax.set_title(tp+' '+net)
im.set_clim([0, 35])
images.append(im)
i+=1
for i in [1,3]:
axs[i].my_remove_axis(xaxis=False, yaxis=True,keep_ticks=True)
ax=axs[1]
im=images[1]
box = ax.get_position()
axColor = pylab.axes([box.x0 + box.width *1.03,
box.y0+box.height*0.1,
0.01,
box.height*0.8])
cbar=pylab.colorbar(im, cax = axColor, orientation="vertical")
cbar.ax.set_ylabel('Rate (Hz)', rotation=270)
from matplotlib import ticker
tick_locator = ticker.MaxNLocator(nbins=4)
cbar.locator = tick_locator
cbar.update_ticks()
cbar.ax.tick_params(direction='in', length=1, width=0.5)
return fig
def get_fig_axs(scale=4):
kw={'n_rows':2,
'n_cols':1,
'w':72/2.54*11.6*scale,
'h':150*scale,
'fontsize':7*scale,
'frame_hight_y':0.5,
'frame_hight_x':0.7,
'title_fontsize':7*scale,
'font_size':7*scale,
'text_fontsize':7*scale,
'linewidth':1.*scale,
'gs_builder':gs_builder}
# kwargs_fig=kwargs.get('kwargs_fig', kw)
from core import plot_settings as ps
fig, axs=ps.get_figure2(**kw)
return fig, axs
def get_fig_axs():
scale = 4
kw = {
'n_rows': 4,
'n_cols': 2,
'w': 72 / 2.54 * 18 * scale,
'h': 300 * scale,
'fontsize': 7 * scale,
'frame_hight_y': 0.5,
'frame_hight_x': 0.7,
'title_fontsize': 7 * scale,
'font_size': 7 * scale,
'text_fontsize': 7 * scale,
'linewidth': 1. * scale,
'gs_builder': gs_builder
}
# kwargs_fig=kwargs.get('kwargs_fig', kw)
from core import plot_settings as ps
fig, axs = ps.get_figure2(**kw)
return fig, axs
'fs':256.*4,
'noverlap':128*4/2},
'oi_min':.5,
'oi_max':1.5,
'oi_upper':1000.,
'oi_fs':256*4,
'keep':['data'],
'compute_performance_name_and_x':lambda x: [x[0],1]
}
return d
fig, axs=ps.get_figure2(n_rows=4,
n_cols=3,
w=int(72/2.54*11.6*(1+1./2+0.2))*scale,
h=int(0.85*72/2.54*11.6*(1+1./2))*scale,
# w=k.get('w',500),
# h=k.get('h',900),
linewidth=1,
fontsize=7*scale,
title_fontsize=7*scale,
gs_builder=gs_builder)
models=['M1', 'M2', 'FS', 'GA', 'GF', 'GI', 'GP', 'ST','SN',
'GI_MS',
'GA_MS',
'GF_MS',
'GI_FS',
'GA_FS',
'GF_FS']
# models=[m for m in models if not ( m in exclude)]
def plot_coher(d, labelsy, labelsx=[], title_name='Slow wave'):
fig, axs = ps.get_figure2(
n_rows=9,
n_cols=12,
#w=700,
w=23 / 56. * 17.6 * 72 / 2.54,
h=23 / 56. * 17.6 * 72 / 2.54,
# h=700,
fontsize=7,
frame_hight_y=0.5,
frame_hight_x=0.7,
title_fontsize=7,
gs_builder=gs_builder,
linewidth=1.)
for ax in axs:
ax.tick_params(direction='in',
length=2,
width=0.5
# top=False, right=False
)
# fig, axs=ps.get_figure(n_rows=4, n_cols=1, w=500.0*0.65*2, h=400.0*0.65*2, fontsize=16,
# frame_hight_y=0.6, frame_hight_x=0.8,
# title_fontsize=20, text_usetex=False)
# nice_labels={'CTX_M1':r'CTX$\to$$MSN_{D1}$ ',
# 'CTX_M2':r'CTX$\to$$MSN_{D2}$ ',
# 'CTX_ST':r'CTX$\to$STN ',
# 'MS_MS':r'MSN$\to$MSN ',
# 'M1':r'$MSN_{D1}$ ',
# 'M2':r'$MSN_{D2}$ ',
# 'GP':r'GPe ',
# 'SN':r'SNr ',
# 'FS_M2':r'FSN$\to$$MSN_{D2}$ ',
# 'GP_ST':r'GPe$\to$STN ',
# 'GP_FS':r'GPe$\to$FSN ',
# 'GP_GP':r'GPe$\to$GPe ',
# 'FS_FS':r'FSN$\to$FSN ',
# 'M1_SN':r'$MSN_{D1}$$\to$SNr ',
# 'ST_GP':r'STN$\to$GPe ',}
from scripts_inhibition.base_effect_conns import nice_labels
groupings = ['Coherence', 'Phase relation']
#
# nice_labels2={'GA_GA':r'TA vs TA',
# 'GI_GA':r'TI vs TA',
# 'GI_GI':r'TI vs TI',
# 'GP_GP':r'GP vs GP'}
for i in range(len(labelsy)):
if labelsy[i] in nice_labels(version=0).keys():
labelsy[i] = nice_labels(version=0)[labelsy[i]]
l0 = []
l2 = []
for key in sorted(d['bar_obj'].keys()):
# if key in ['GA_GA', 'GI_GA', 'GP_GP']:
# d['bar_obj'][key].y[1,:]+=1
l0.append(d['bar_obj'][key].y[0, :])
l2.append(d['bar_obj'][key].y[1, :])
labelsx.append(key)
z = numpy.transpose(numpy.array(l0 + l2))
for i in range(len(labelsx)):
if labelsx[i] in nice_labels(version=1).keys():
labelsx[i] = nice_labels(version=1)[labelsx[i]]
_vmin = 0
_vmax = 4
stepx = 1
stepy = 1
startx = 0
starty = 0
stopy = 14
stopx = 8
maxy = 14
maxx = 8
posy = numpy.linspace(0.5, maxy - 0.5, maxy)
posx = numpy.linspace(0.5, maxx - 0.5, maxx)
axs[1].barh(posy,
numpy.mean(z, axis=1)[::-1],
align='center',
color='0.5',
edgecolor='0.5')
axs[1].plot([1, 1], [0, stopy], 'k', linestyle='--')
x1, y1 = numpy.meshgrid(numpy.linspace(startx, stopx, maxx + 1),
numpy.linspace(stopy, starty, maxy + 1))
im = axs[0].pcolor(x1, y1, z, cmap='jet', vmin=_vmin, vmax=_vmax)
axs[0].set_yticks(posy)
axs[0].set_yticklabels(labelsy[::-1])
axs[0].set_xticks(posx)
axs[0].set_xticklabels(labelsx * 2, rotation=70, ha='right')
axs[0].set_ylim([0, maxy])
axs[0].text(0.25,
-0.39,
"Coherence",
ha='center',
transform=axs[0].transAxes)
axs[0].text(0.75,
-0.39,
"Phase shift",
ha='center',
transform=axs[0].transAxes)
axs[1].text(0.5,
-0.15,
"Mean",
transform=axs[1].transAxes,
ha='center',
rotation=0)
axs[1].text(0.5,
-0.22,
"effect",
transform=axs[1].transAxes,
ha='center',
rotation=0)
font0 = FontProperties()
font0.set_weight('bold')
axs[1].text(
1.3,
0.5,
title_name,
# fontsize=28,
va='center',
transform=axs[0].transAxes,
rotation=270,
fontproperties=font0)
axs[0].text(-0.58,
0.5,
"Connection without dop. effect",
transform=axs[0].transAxes,
rotation=90,
ha='center',
va='center')
axs[1].my_remove_axis(xaxis=False, yaxis=True)
axs[1].my_set_no_ticks(xticks=2)
axs[1].set_xlim([0, maxy])
axs[1].set_xlim([0, 2])
# axs[1].set_xticks([0.04, 0.12])
box = axs[0].get_position()
axColor = pylab.axes([
box.x0 + 0.0 * box.width, box.y0 + box.height + box.height * 0.15,
box.width * 0.8, 0.02
])
# axColor = pylab.axes([0.05, 0.9, 1.0, 0.05])
cbar = pylab.colorbar(im, cax=axColor, orientation="horizontal")
cbar.ax.set_title('Deviation from base model MSE') #, rotation=270)
cbar.ax.tick_params(direction='in',
length=1,
width=0.5
# top=False, right=False
)
from matplotlib import ticker
tick_locator = ticker.MaxNLocator(nbins=4)
cbar.locator = tick_locator
cbar.update_ticks()
return fig
#
#
#
#
# # show_heat_map(dd, 'mean_rate_slices', **k)
# # show_variability(dd, 'mean_rate_slices', **k)
# # show_variability(dd, 'mean_rate_slices', net='Net_01',**k)
# show_variability(dd, 'mean_rate_slices', net='Net_03',**k)
# show_variability(dd, 'mean_rate_slices', net='Net_05',**k)
# show_variability(dd, 'mean_rate_slices', net='Net_06',**k)
scale=4
fig, axs=ps.get_figure2(n_rows=4,
n_cols=2,
w=int(72/2.54*8.9) *scale,
h=450*scale,
fontsize=7*scale,
title_fontsize=7*scale,
# grid=[3,2],
gs_builder=gs_builder_new)
print len(axs)
# pylab.show()
k={'axs':axs,
# 'do_colorbar':False,
'fig':fig,
'model':'SN',
'print_statistics':False,
'resolution':10,
def create_figs(setup, file_name_figs, d, models):
sd_figs = Storage_dic.load(file_name_figs)
figs = []
# d_plot_fr = setup.plot_fr()
d_plot_fr2=setup.plot_fr2()
# d_plot_mr=setup.plot_mr()
d_plot_mr_diff=setup.plot_mr_diff()
fig, axs=ps.get_figure2(**setup.plot_fig_axs().get('fig_and_axes'))
figs.append(fig)
for ax in axs:
ax.tick_params(direction='out',
length=2,
width=0.5,
pad=0.01,
top=False, right=False
)
# for name in sorted(d.keys()):
# if name=='Difference':
# continue
# for model in d[name]['set_0'].keys():
# v={'firing_rate':d[name]['set_0'][model]['firing_rate']}
# d['Net_0'][model]=v
# figs.append(show_fr(d, models, **d_plot_fr))
show_fr(d, models, axs[0:2], **d_plot_fr2)
axs=figs[-1].get_axes()
for i, ax in enumerate(axs[2:4]):
if i==0:
ax.set_ylim([0,70])
else:
ax.set_ylim([0,70])
ax.my_set_no_ticks(xticks=4)
show_mr_diff(d, models, axs[2:4], **d_plot_mr_diff)
# axs=figs[-1].get_axes()
for i, ax in enumerate(axs[2:4]):
handles, labels=ax.get_legend_handles_labels()
ax.legend(handles, labels,loc='upper left')
for i, s, c0, c1, rotation in [
# [2, 'Ralative', -0.58, 0., 90],
# [2, 'decrease (Hz)', -0.39, 0., 90],
[0, 'Firing rate (Hz)', -0.45, 0., 90],
[0, r'$MSN_{D1}$', -0.27, 0.5, 90],
[1, r'$MSN_{D2}$', -0.27, 0.5, 90],
# [2, r'$MSN_{D1}$', 1.1, 0.5, 270],
# [3, r'$MSN_{D2}$', 1.1, 0.5, 270],
]:
axs[i].text(c0, c1, s,
fontsize=7,
transform=axs[i].transAxes,
verticalalignment='center',
horizontalalignment='center',
rotation=rotation)
for i, ax in enumerate(axs[1:4]):
ax.legend().set_visible(False)
axs[0].legend(axs[0].lines[:],
d_plot_fr2['labels'][:],
bbox_to_anchor=(1.2, 1.8), ncol=1,
handletextpad=0.1,
frameon=False,
columnspacing=0.3,
labelspacing=0.2)
axs[0].my_remove_axis(xaxis=True, yaxis=False,
keep_ticks=False)
axs[2].my_remove_axis(xaxis=True, yaxis=False,
keep_ticks=False)
for i, ax in enumerate(axs):
ax.set_ylabel('')
ax.my_set_no_ticks(yticks=3, xticks=3)
if i==0:
# ax.set_xticks([700, 1000, 1300])
ax.set_ylim([0,60])
ax.set_xlim([700,1600])
if i==1:
ax.set_xticks([700, 1000, 1300])
ax.set_xticklabels([0, 300, 600])
ax.set_ylim([0,50])
ax.set_xlim([700,1600])
if i==2:
ax.set_title('Difference')
# ax.set_yticks([0, 1000.0])
ax.set_ylim([0,30])
if i==3:
# ax.set_yticks([0, 1000.0])
ax.set_ylim([0,25])
sd_figs.save_figs(figs, format='png', dpi=400)
sd_figs.save_figs(figs, format='svg', in_folder='svg')
print name, net
if not (net in d[name].keys()):
i+=1
continue
dd['Net_{:0>2}'.format(i)]=d[name][net]
titles.append(name+'_'+net)
i+=1
pp(dd)
val=int(72/2.54*17.6*(1-17./48))
scale=1
fig, axs=ps.get_figure2(n_rows=11,
n_cols=11,
w=val*scale,
h=300*scale,
fontsize=7*scale,
title_fontsize=7*scale,
gs_builder=gs_builder)
k={'axs':axs,
'do_colorbar':False,
'fig':fig,
'models':['SN'],
'print_statistics':False,
'resolution':10,
'titles':['']*5*5,
'type_of_plot':'mean',
'vlim_rate':[-100, 100],
'marker_size':8}
s = 'Net_{:0>2}'.format(i)
dd[s] = d[name][net]
translation[s] = name + '_' + net
# titles.append(name+'_'+net)
i += 1
pp(dd)
pp(translation)
scale = 1
figs = []
fun_call = [show_heat_map, show_variability_several]
for iFig in range(2):
fig, axs = ps.get_figure2(n_rows=9,
n_cols=10,
w=72 / 2.54 * 17.6 * scale,
h=72 / 2.54 * 17.6 * scale * 1.1,
fontsize=7 * scale,
title_fontsize=7 * scale,
gs_builder=gs_builder)
for ax in axs:
ax.tick_params(
direction='in',
length=0,
width=0.5,
# pad=1,
top=False,
right=False,
left=False,
bottom=False,
)
def plot_coher(d, labelsy, labelsx=[], title_name='Slow wave'):
fig, axs=ps.get_figure2(n_rows=9, n_cols=12,
#w=700,
w=23/56.*17.6*72/2.54,
h=23/56.*17.6*72/2.54,
# h=700,
fontsize=7,
frame_hight_y=0.5, frame_hight_x=0.7,
title_fontsize=7,
gs_builder=gs_builder,
linewidth=1.)
for ax in axs:
ax.tick_params(direction='in',
length=2,
width=0.5
# top=False, right=False
)
# fig, axs=ps.get_figure(n_rows=4, n_cols=1, w=500.0*0.65*2, h=400.0*0.65*2, fontsize=16,
# frame_hight_y=0.6, frame_hight_x=0.8,
# title_fontsize=20, text_usetex=False)
# nice_labels={'CTX_M1':r'CTX$\to$$MSN_{D1}$ ',
# 'CTX_M2':r'CTX$\to$$MSN_{D2}$ ',
# 'CTX_ST':r'CTX$\to$STN ',
# 'MS_MS':r'MSN$\to$MSN ',
# 'M1':r'$MSN_{D1}$ ',
# 'M2':r'$MSN_{D2}$ ',
# 'GP':r'GPe ',
# 'SN':r'SNr ',
# 'FS_M2':r'FSN$\to$$MSN_{D2}$ ',
# 'GP_ST':r'GPe$\to$STN ',
# 'GP_FS':r'GPe$\to$FSN ',
# 'GP_GP':r'GPe$\to$GPe ',
# 'FS_FS':r'FSN$\to$FSN ',
# 'M1_SN':r'$MSN_{D1}$$\to$SNr ',
# 'ST_GP':r'STN$\to$GPe ',}
from scripts_inhibition.base_effect_conns import nice_labels
groupings=['Coherence','Phase relation']
#
# nice_labels2={'GA_GA':r'TA vs TA',
# 'GI_GA':r'TI vs TA',
# 'GI_GI':r'TI vs TI',
# 'GP_GP':r'GP vs GP'}
for i in range(len(labelsy)):
if labelsy[i] in nice_labels(version=0).keys():
labelsy[i]=nice_labels(version=0)[labelsy[i]]
l0=[]
l2=[]
for key in sorted(d['bar_obj'].keys()):
# if key in ['GA_GA', 'GI_GA', 'GP_GP']:
# d['bar_obj'][key].y[1,:]+=1
l0.append(d['bar_obj'][key].y[0,:])
l2.append(d['bar_obj'][key].y[1,:])
labelsx.append(key)
z=numpy.transpose(numpy.array(l0+l2))
for i in range(len(labelsx)):
if labelsx[i] in nice_labels(version=1).keys():
labelsx[i]=nice_labels(version=1)[labelsx[i]]
_vmin=0
_vmax=4
stepx=1
stepy=1
startx=0
starty=0
stopy=14
stopx=8
maxy=14
maxx=8
posy=numpy.linspace(0.5,maxy-0.5, maxy)
posx=numpy.linspace(0.5,maxx-0.5, maxx)
axs[1].barh(posy,numpy.mean(z,axis=1)[::-1],
align='center',
color='0.5',
edgecolor='0.5'
)
axs[1].plot([1,1], [0,stopy], 'k', linestyle='--')
x1,y1=numpy.meshgrid(numpy.linspace(startx, stopx, maxx+1),
numpy.linspace(stopy, starty, maxy+1))
im = axs[0].pcolor(x1, y1, z, cmap='jet',
vmin=_vmin, vmax=_vmax
)
axs[0].set_yticks(posy)
axs[0].set_yticklabels(labelsy[::-1])
axs[0].set_xticks(posx)
axs[0].set_xticklabels(labelsx*2, rotation=70, ha='right')
axs[0].set_ylim([0,maxy])
axs[0].text(0.25, -0.39, "Coherence",
ha='center',
transform=axs[0].transAxes)
axs[0].text(0.75, -0.39, "Phase shift",
ha='center', transform=axs[0].transAxes)
axs[1].text(0.5, -0.15, "Mean",
transform=axs[1].transAxes,
ha='center',
rotation=0)
axs[1].text(0.5, -0.22, "effect",
transform=axs[1].transAxes,
ha='center',
rotation=0)
font0 = FontProperties()
font0.set_weight('bold')
axs[1].text(1.3, 0.5, title_name,
# fontsize=28,
va='center',
transform=axs[0].transAxes,
rotation=270,
fontproperties=font0)
axs[0].text(-0.58, 0.5, "Connection without dop. effect", transform=axs[0].transAxes,
rotation=90,
ha='center',
va='center')
axs[1].my_remove_axis(xaxis=False, yaxis=True)
axs[1].my_set_no_ticks(xticks=2)
axs[1].set_xlim([0,maxy])
axs[1].set_xlim([0,2])
# axs[1].set_xticks([0.04, 0.12])
box = axs[0].get_position()
axColor=pylab.axes([box.x0+0.0*box.width,
box.y0+box.height+box.height*0.15,
box.width*0.8,
0.02])
# axColor = pylab.axes([0.05, 0.9, 1.0, 0.05])
cbar=pylab.colorbar(im, cax = axColor, orientation="horizontal")
cbar.ax.set_title('Deviation from base model MSE')#, rotation=270)
cbar.ax.tick_params(direction='in',
length=1,
width=0.5
# top=False, right=False
)
from matplotlib import ticker
tick_locator = ticker.MaxNLocator(nbins=4)
cbar.locator = tick_locator
cbar.update_ticks()
return fig
def create_figs(file_name_figs, from_disks, d, models, setup):
sd_figs = Storage_dic.load(file_name_figs)
# d_plot_fr = setup.plot_fr()
d_plot_mr = setup.plot_mr()
d_plot_mr2 = setup.plot_mr2()
figs = []
pp(setup.plot_mr_general().get('fig_and_axes'))
fig, axs=ps.get_figure2(**setup.plot_mr_general().get('fig_and_axes'))
figs.append(fig)
for ax in axs:
ax.tick_params(direction='out',
length=2,
width=0.5,
pad=0.01,
top=False, right=False
)
show_mr(d, ['M1', 'M2'], axs[2:4], **d_plot_mr2)
for i, s, c0, c1, rotation in [[2, 'Rel. inh. effect', -0.31, 0., 90],
# [2, 'decrease (Hz)', -0.39, 0., 90],
[0, 'Firing rate (Hz)', -0.45, 0., 90],
[0, r'$MSN_{D1}$', -0.27, 0.5, 90],
[1, r'$MSN_{D2}$', -0.27, 0.5, 90],
# [2, r'$MSN_{D1}$', 1.1, 0.5, 270],
# [3, r'$MSN_{D2}$', 1.1, 0.5, 270],
]:
axs[i].text(c0, c1, s,
fontsize=7,
transform=axs[i].transAxes,
verticalalignment='center',
horizontalalignment='center',
rotation=rotation)
# ps.shift('left', axs, 0.5, n_rows=len(axs), n_cols=1)
for ax in axs[2:4]:
if not ax.legend():
continue
# ax.legend(bbox_to_anchor=(2.2, 1))
ax.legend().set_visible(False)
# ax.set_ylabel('')
# ax.set_xlim([1,2])
# figs.append(show_fr(d, models, **d_plot_fr))
# figs.append(show_mr(d, models, **d_plot_mr))
# figs.append(show_mr(d, models, **d_plot_mr2))
show_mr(d, ['M1', 'M2'], axs[0:2], **d_plot_mr)
# axs=figs[-1].get_axes()
# ps.shift('left', axs, 0.5, n_rows=len(axs), n_cols=1)
axs[0].legend(axs[0].lines[0:6],
d_plot_mr2['labels'][0:6],
bbox_to_anchor=(2.4, 2.4), ncol=2,
handletextpad=0.1,
frameon=False,
columnspacing=0.3,
labelspacing=0.2)
for i, ax in enumerate(axs[1:2]):
if ax.legend():
ax.legend().set_visible(False)
# ax.legend(bbox_to_anchor=(2.2, 1))
# ax.set_xlim([1,1.5])
if i==1:
ax.set_ylim([0,20])
axs[0].my_remove_axis(xaxis=True, yaxis=False,
keep_ticks=False)
axs[2].my_remove_axis(xaxis=True, yaxis=False,
keep_ticks=False)
for i, ax in enumerate(axs):
ax.set_ylabel('')
ax.my_set_no_ticks(yticks=3, xticks=4)
# ax.set_xticks([1.1,1.3,1.5])
if i==0:
ax.set_yticks([0,10,20])
if i==1:
ax.set_yticks([0,7,14])
ax.set_ylim([0,20])
if len(axs[2].lines)>2:
axs[2].lines.remove(axs[2].lines[0])
axs[2].lines.remove(axs[2].lines[-1])
sd_figs.save_figs(figs, format='png', dpi=100)
sd_figs.save_figs(figs, format='svg', in_folder='svg')
def show_heat_map(d, attr, **k):
do_colorbar=k.get('do_colorbar',True)
models=['SN']
print_statistics=k.get('print_statistics', True)
res=k.get('resolution')
titles=k.get('titles')
vlim_variance=k.get('vlim_variance')
vlim_CV=k.get('vlim_CV')
vlim_rate=k.get('vlim_rate')
axs=k.get('axs')
fig=k.get('fig')
if not axs or not fig:
fig, axs=ps.get_figure2(n_rows=3,
n_cols=8,
w=780/(24./7.),
h=910/(24./7.),
fontsize=7,
title_fontsize=7,
gs_builder=gs_builder)
k['fig']=fig
k['axs']=axs
type_of_plot=k.get('type_of_plot', 'mean')
i=0
performance={}
m=len(d.keys())
for model in models:
for key in sorted(d.keys()):
print key
i=int(key.split('_')[1])
obj0=d[key]['set_0'][model][attr]
obj1=d[key]['set_1'][model][attr]
args=[obj0.x_set, obj1.x_set,
numpy.mean(obj1.y_raw_data-obj0.y_raw_data, axis=0),
numpy.std(obj1.y_raw_data-obj0.y_raw_data, axis=0),
numpy.mean(obj0.y_raw_data, axis=0),
numpy.mean(obj1.y_raw_data, axis=0)]
for j, arg in enumerate(args):
arg.shape
args[j]=numpy.reshape(arg, [res,res])
x, y, z, z_std, d0, d1=args
if type_of_plot=='variance':
z=z_std
_vmin, _vmax=vlim_variance
elif type_of_plot=='CV':
z=z_std/numpy.abs(z)
_vmin, _vmax=vlim_CV
else:
_vmin, _vmax=vlim_rate
stepx=(x[0,-1]-x[0,0])/res
stepy=(y[-1,0]-y[0,0])/res
x1,y1=numpy.meshgrid(numpy.linspace(x[0,0], x[0,-1], res+1),
numpy.linspace(y[0,0], y[-1,0], res+1))
x2,y2=numpy.meshgrid(numpy.linspace(x[0,0]+stepx/2,
x[0,-1]-stepx/2, res),
numpy.linspace(y[0,0]+stepy/2,
y[-1,0]-stepy/2, res))
thr=k.get('threshold',14)
im = axs[i].pcolor(x1, y1, z, cmap='coolwarm',
vmin=_vmin, vmax=_vmax)
for m in ['d', 's', 'n']:
set_action_selection_marker(i, d0, d1,
x2, y2, thr,
marker=m,
**k)
performance[key] = numpy.round(numpy.sum(numpy.abs(z)), 2)
box = axs[i].get_position()
axs[i].set_xlabel('Action 1')
axs[i].set_ylabel('Action 2')
axs[i].set_xlim([x[0,0], x[0,-1]])
axs[i].set_ylim([y[0,0], y[-1,0]])
# create color bar
if key=='Net_0' and do_colorbar:
label='Contrast (spike/s)'
box = axs[0].get_position()
axColor = pylab.axes([box.x0 + box.width * 2.56,
box.y0+box.height*0.1,
0.02,
box.height*0.8])
cbar=pylab.colorbar(im, cax = axColor, orientation="vertical")
cbar.ax.set_ylabel(label, rotation=270)
cbar.set_ticks([_vmin,_vmax])
# from matplotlib import ticker
#
# tick_locator = ticker.MaxNLocator(nbins=3)
# cbar.locator = tick_locator
# cbar.update_ticks()
axs[i].text(0.5, k.get('pos_ax_titles',1.05) , titles[i],
horizontalalignment='center',
transform=axs[i].transAxes,
fontsize=k.get('fontsize_ax_titles',7))
# axs[i].set_title(titles[i])
# i+=1
if print_statistics:
ax=axs[-1]
import pprint
ax.text( 0.1, 0.1, pprint.pformat(performance),
transform=ax.transAxes,
fontsize=7)
return fig, performance
def show_rate_D1_D2_SNR(d, d_plot_fr2):
fig, axs=ps.get_figure2(**d_plot_fr2.get('fig_and_axes'))
pp(d)
show_fr(d, ['M1', 'M2', 'SN'], axs, **d_plot_fr2)
for i, s, c0, c1, rotation in [
[1, 'Firing rate (Hz)', -0.45, 0.5, 90],
[0, r'$D1$', -0.25, 0.5, 90],
[1, r'$D2$', -0.25, 0.5, 90],
[2, r'SNr', -0.25, 0.5, 90],
]:
axs[i].text(c0, c1, s,
fontsize=7,
transform=axs[i].transAxes,
verticalalignment='center',
horizontalalignment='center',
rotation=rotation)
axs[0].legend(axs[0].lines[0:6],
['Action 1', 'Action 2'],
bbox_to_anchor=(1.15, 1.85),
ncol=2,
handletextpad=0.1,
frameon=False,
columnspacing=0.3,
labelspacing=0.2)
for i, ax in enumerate(axs):
ax.my_set_no_ticks(xticks=3, yticks=2)
if i==2:
ax.set_yticks([0,60])
ax.set_xlabel('')
ax.text(0.5,-0.65,'Time (ms)',
fontsize=7,
transform =ax.transAxes,
ha='center',va='center')
if i==1:
ax.set_yticks([0,20])
#
if i==0:
ax.set_yticks([0,30])#
ax.set_ylabel('')
axs[0].my_remove_axis(xaxis=True, yaxis=False,
keep_ticks=True)
axs[1].my_remove_axis(xaxis=True, yaxis=False,
keep_ticks=True)
for ax in axs:
ax.tick_params(direction='out',
length=1,
width=0.5,
top=False, right=False
)
#think you have to to this after toget it to work
for ax in axs:
ax.tick_params(
pad=1,
)
for i, ax in enumerate(axs[1:3]):
if not ax.legend():
continue
ax.legend().set_visible(False)
return fig
def create_figs(setup, file_name_figs, d, models):
sd_figs = Storage_dic.load(file_name_figs)
figs = []
d_plot_3d=setup.plot_3d()
d_plot_fr = setup.plot_fr()
d_plot_fr2 = setup.plot_fr2()
d_plot_bulk=setup.plot_bulkt()
l_mean_rate_slices=setup.get_l_mean_rate_slices()
figs.append(plot_optimal(setup.res))
fig, axs=ps.get_figure2(n_rows=3,
n_cols=8,
w=780/(24./7.),
h=910/(24./7.),
fontsize=7,
title_fontsize=7,
gs_builder=gs_builder)
pp(d)
for name in l_mean_rate_slices:
d_plot_3d['type_of_plot']='mean'
fig, perf=show_heat_map(d, name, **d_plot_3d)
figs.append(fig)
for name in l_mean_rate_slices:
d_plot_3d['type_of_plot']='variance'
fig, perf=show_heat_map(d, name, **d_plot_3d)
figs.append(fig)
for name in l_mean_rate_slices:
d_plot_3d['type_of_plot']='CV'
fig, perf=show_heat_map(d, name, **d_plot_3d)
figs.append(fig)
for name in l_mean_rate_slices:
figs.append(show_bulk(d, models, name, **d_plot_bulk))
# pylab.show()
for i in range(len(d.keys())):
if i!=1 and 1<len(d.keys()):
continue
try:figs.append(show_rate_D1_D2_SNR(d['Net_'+str(i)], d_plot_fr2))
except: warnings.warn('failed rate plot')
try:figs.append(show_rate_all(d['Net_'+str(i)], d_plot_fr))
except: warnings.warn('failed rate plot')
# figs.append()
sd_figs.save_figs(figs, format='png', dpi=400)
sd_figs.save_figs(figs, format='svg', in_folder='svg')
'add_midpoint':False,
'psd':{'NFFT':128*4,
'fs':256.*4,
'noverlap':128*4/2},
'oi_min':.5,
'oi_max':1.5,
'oi_upper':1000.,
'oi_fs':256*4,
'keep':['data']}
return d
fig, axs=ps.get_figure2(n_rows=4,
n_cols=3,
w=int(72/2.54*11.6*(1+1./2+0.2))*scale,
h=int(0.85*72/2.54*11.6*(1+1./2))*scale,
# w=k.get('w',500),
# h=k.get('h',900),
linewidth=1,
fontsize=7*scale,
title_fontsize=7*scale,
gs_builder=gs_builder)
models=['M1', 'M2', 'FS', 'GA', 'GI', 'GP', 'ST','SN',
'GI_ST', 'GP_GP', 'GA_GA', 'GI_GA', 'GI_GI']
# models=[m for m in models if not ( m in exclude)]
nets=['Net_0', 'Net_1']
attrs=[
'firing_rate',
'mean_coherence',
'phases_diff_with_cohere',
dd['Net_{:0>2}'.format(i)] = d[name][net]
# titles.append(name+'_'+net)
i += 1
# for name, nets in builder:
# for net in nets:
# dd['Net_{}'.format(i)]=d[name][net]
# i+=1
pp(dd)
print len(dd['Net_00']['set_1']['SN']['mean_rate_slices'].y)
figs = []
fig, axs = ps.get_figure2(n_rows=19,
n_cols=16,
w=int(72 / 2.54 * 17.6 * (17. / 48)),
h=300,
fontsize=7,
title_fontsize=7,
gs_builder=gs_builder)
figs.append(fig)
k = {
'axs':
axs,
'do_colorbar':
False,
'fig':
fig,
'models': ['SN'],
'print_statistics':
False,
'resolution':
dd['Net_{:0>2}'.format(i)]=d[name][net]
# titles.append(name+'_'+net)
i+=1
# for name, nets in builder:
# for net in nets:
# dd['Net_{}'.format(i)]=d[name][net]
# i+=1
pp(dd)
print len(dd['Net_00']['set_1']['SN']['mean_rate_slices'].y)
figs=[]
fig, axs=ps.get_figure2(n_rows=19,
n_cols=16,
w=int(72/2.54*17.6*(17./48)),
h=300,
fontsize=7,
title_fontsize=7,
gs_builder=gs_builder)
figs.append(fig)
k={'axs':axs,
'do_colorbar':False,
'fig':fig,
'models':['SN'],
'print_statistics':False,
'resolution':7,
'threshold':thr,
'titles':['Only D1',
'D1 & D2',
r'No MSN$\to$MSN',
r'No FSN$\to$MSN',
n=10
z=numpy.random.random((n,n))
z[1,:]+=1
z[8,:]+=2
_vmin=0
_vmax=4
stepx=1
stepy=1
startx=0
starty=0
stopy=10
stopx=10
res=10
fig, axs=ps.get_figure2(n_rows=5, n_cols=6, w=700, h=500, fontsize=24,
frame_hight_y=0.5, frame_hight_x=0.7, title_fontsize=20,
gs_builder=gs_builder)
pos=numpy.linspace(0.5,9.5,10)
axs[1].barh(pos,numpy.mean(z,axis=1)[::-1], align='center')
# ax=pylab.subplot(111)
nets=['Net_'+str(i) for i in range(10)]
x1,y1=numpy.meshgrid(numpy.linspace(startx, stopx, res+1),
numpy.linspace(stopy, starty, res+1))
# x2,y2=numpy.meshgrid(numpy.linspace(startx+stepx/2,
# stopx-stepx/2, res),
# numpy.linspace(stopy+stepy/2,
# starty-stepy/2, res))
