def _initGP(self,colCovarType,rank_r,rank_g,rank_n):
"""
Initializes genetic and noise LIMIX trait covar
Args:
colCovarType_g: genetic trait covariance type
colCovarType_n: noise trait covariance type
rank_r: rank of region trait covar
rank_g: rank of genetic trait covar
rank_n: rank of noise trait covar
"""
if self.P==1:
self.rank_r = 1
self.rank_g = 1
self.rank_n = 1
# ideally should be diag
colCovarType = 'freeform'
elif colCovarType is None:
colCovarType='freeform'
self.rank_r = rank_r
self.rank_g = rank_g
self.rank_n = rank_n
self.colCovarType = colCovarType
self.Cg = self._buildTraitCovar(colCovarType,rank_g)
self.Cn = self._buildTraitCovar(colCovarType,rank_n)
XXnotNone = self.XX is not None
SUnotNone = self.S_XX is not None and self.U_XX is not None
# build mean
if self.bgRE:
self.gp = gp3kronSum(self.mean,self.Cg,self.Cn,XX=self.XX,S_XX=self.S_XX,U_XX=self.U_XX,rank=self.rank_r)
else:
self.gp = gp2kronSumLR(self.Y,self.Cn,F=self.F,rank=self.rank_r)
def _initGP(self, colCovarType, rank_r, rank_g, rank_n):
"""
Initializes genetic and noise LIMIX trait covar
Args:
colCovarType_g: genetic trait covariance type
colCovarType_n: noise trait covariance type
rank_r: rank of region trait covar
rank_g: rank of genetic trait covar
rank_n: rank of noise trait covar
"""
if self.P == 1:
self.rank_r = 1
self.rank_g = 1
self.rank_n = 1
# ideally should be diag
colCovarType = 'freeform'
elif colCovarType is None:
colCovarType = 'freeform'
self.rank_r = rank_r
self.rank_g = rank_g
self.rank_n = rank_n
self.colCovarType = colCovarType
self.Cg = self._buildTraitCovar(colCovarType, rank_g)
self.Cn = self._buildTraitCovar(colCovarType, rank_n)
XXnotNone = self.XX is not None
SUnotNone = self.S_XX is not None and self.U_XX is not None
# build mean
if self.bgRE:
self.gp = gp3kronSum(self.mean,
self.Cg,
self.Cn,
XX=self.XX,
S_XX=self.S_XX,
U_XX=self.U_XX,
rank=self.rank_r)
else:
self.gp = gp2kronSumLR(self.Y, self.Cn, F=self.F, rank=self.rank_r)
gp.setParams(params)
gp.checkGradient()
print "test optimization"
conv,info = OPT.opt_hyper(gp,params,factr=1e3)
print conv
ipdb.set_trace()
if 1:
# generate parameters
params = {}
params['Cr'] = SP.randn(P)
params['Cg'] = SP.randn(int(0.5*P*(P+1)))
params['Cn'] = SP.randn(int(0.5*P*(P+1)))
params['mean'] = 1e-2*SP.randn(mean.getParams().shape[0])
print "check gradient with gp3kronSum"
gp = gp3kronSum(mean,Cg,Cn,XX,Xr=Xr)
gp.setParams(params)
gp.LMLgrad()
gp.checkGradient()
print "test optimization"
conv,info = OPT.opt_hyper(gp,params,factr=1e3)
ipdb.set_trace()
m,cov,M,S = gp.getPosteriorSnpWeights(matrix=True)
print conv
ipdb.set_trace()
if 1:
# generate parameters
params = {}
params['Cr'] = SP.randn(P)
params['Cn'] = SP.randn(int(0.5*P*(P+1)))