def _test_LNP( rgc_type='off parasol' ):
vardict = LNP( **thetaM( **linear_reparameterization()))
init_LNP = LNP_model( init_sv1(rgc_type) )
indices = extract( linear_stats( rgc_type, (5,0) ), ['sparse_index', 'subunit_index'] )
indices['N_subunits'] = len(cones)
unknown = extract(init_LNP,['sv1'])
train_LNP = global_objective( unknown, extract(init_LNP,['u','V2']),
vardict, run=linear_stats( rgc_type, (5,0) ),
indices=indices)
train_LNP.with_callback(callback)
train_LNP.description = 'LNP'
sv1 = optimize.optimizer( train_LNP )( init_params=unknown, maxiter=5000, gtol=1e-7 )
model = LNP_model( train_LNP.unflat( sv1 )['sv1'] )
model['LL'] = global_objective( unknown, extract(init_LNP,['u','V2']),
vardict, run=linear_stats( rgc_type, (-5,0)),
indices=indices).LL(sv1)
save(model,'LNP_'+rgc_type)
return model
def callback( objective , params , force=False , other={} , objectives=[] , plot_every=42):
result = objective.unflat(params)
# ipdb.set_trace()
result.update(other)
for o in objectives:
result.update( {o.description:o.LL(params)} )
for name,value in result.items():
if hasattr(value, '__call__'):
result[name] = value(params)
try:
display_params( result )
except:
pass
if force or numpy.remainder( iterations[0] , plot_every ) == plot_every-1:
filename = objective.description+('iter%d'%iterations[0])
# filename = objective.description+'_'+'_'.join(sorted(result.keys()))+'_'+ \
# rgc_type+'_lam'+str(int(100*lam))+'_'+str(int(maxiter))+'iters'
save(result,filename)
plot_params( result, filename )
iterations[0] = iterations[0] + 1
def ARD( stats , lam=0.00 ):
stats['D'], stats['Z'] = schur(stats['cov'])
print 'Starting ARD of size ', stats['Z'].shape,' with lambda=',lam
sys.stdout.flush()
D,Z = stats['D']/2 , stats['Z']
print 'Schur decomposition completed'
sys.stdout.flush()
DD = numpy.diag(D)
keep= DD>1e-10
P = (Z[:,keep] * numpy.sqrt(DD[keep])).T
dSTA = numpy.concatenate(
[STA[:,numpy.newaxis]-stats['mean'][:,numpy.newaxis]
for STA in stats['STA']], axis=1)
y = numpy.dot ( (Z[:,keep] * 1/numpy.sqrt(DD[keep])).T , dSTA ) / 2
iW = 1e-1
for i in range(1):
print 'Irlsing'
sys.stdout.flush()
V, iW = IRLS.IRLS( y, P, x=0, disp_every=1, lam=lam, maxiter=2 ,
ftol=1e-5, nonzero=1e-1, iw=iW)
save({'V':V,'iW':iW},'Localizing_lam%.0e'%lam)
return V, iW
# print
# print
# print
# print rgctype
# print
# print
# infile = 're2STD_Uc2_'+rgctype
#
# indices = extract( linear_stats(rgctype,(5,0)), ['sparse_index', 'subunit_index'] )
# indices['N_subunits'] = len(cones)
# iterations[0] = 0
# retrain = optimize_LQLEP(rgctype, filename=infile, indices=indices,
# description='posV1_L2v1_smooth10_c_', maxiter=42*5+2,
# vardict = LQLEP_positiveV1( **LQLEP_wBarrier( **LQLEP(
# **thetaM( **u2c_parameterization())))))
for i in range(41,214,42)+[214]:
filename = ('posV1_L2v1_newmooth_c_off parasoliter%d'%i)
rgctype = 'off parasol'
# plot_thetaM( filename, rgctype )
e = exact_normalized_LLs( filename, rgctype )
save( e, filename+'_LL' )
print
print 'Calculating simulated_STAC for', filename
s = simulated_STAC( filename, rgctype)
#for rgctype in types:
# infile = 're2STD_usmooth_'
# e = exact_LL( filename=infile, rgctype=rgctype )
# save( e, infile+'_LL' )
