def __init__(self,
kinship,
design,
kinship_cm=None,
kinship_cross=None,
jitter=1e-4):
if kinship_cm is None: kinship_cm = kinship
if kinship_cross is None: kinship_cross = kinship
ff_dim = 2
self.covff = FreeFormCov(ff_dim, jitter=1e-4)
self._K = kinship
self._ZK = sp.dot(design, kinship_cross.T)
self._KZ = sp.dot(kinship_cross, design.T)
self._ZKZ = sp.dot(design, sp.dot(kinship_cm, design.T))
Covariance.__init__(self, kinship.shape[0])
def __init__(self, kinship, design, kinship_cm = None, kinship_cross = None, jitter = 1e-4):
if kinship_cm is None: kinship_cm = kinship
if kinship_cross is None: kinship_cross = kinship
ff_dim = 2
self.covff = FreeFormCov(ff_dim, jitter = 1e-4)
self._K = kinship
self._ZK = sp.dot(design, kinship_cross.T)
self._KZ = sp.dot(kinship_cross, design.T)
self._ZKZ = sp.dot(design, sp.dot(kinship_cm, design.T))
Covariance.__init__(self, kinship.shape[0])
class DirIndirCov(Covariance):
"""
Covariance matrix for decomposing direct and social genetic effects
"""
def __init__(self,
kinship,
design,
kinship_cm=None,
kinship_cross=None,
jitter=1e-4):
if kinship_cm is None: kinship_cm = kinship
if kinship_cross is None: kinship_cross = kinship
ff_dim = 2
self.covff = FreeFormCov(ff_dim, jitter=1e-4)
self._K = kinship
self._ZK = sp.dot(design, kinship_cross.T)
self._KZ = sp.dot(kinship_cross, design.T)
self._ZKZ = sp.dot(design, sp.dot(kinship_cm, design.T))
Covariance.__init__(self, kinship.shape[0])
def dirIndirCov_K(self):
return self.covff.K()
def dirIndirCov_K_ste(self):
return self.covff.K_ste()
#####################
# Properties
#####################
@property
def variance(self):
return self.covff.variance
@property
def correlation(self):
return self.covff.correlation
#####################
# Params handling
#####################
def getParams(self):
return self.covff.getParams()
def setParams(self, params):
self.covff.setParams(params)
self.clear_all()
def getNumberParams(self):
return self.covff.getNumberParams()
def setCovariance(self, cov):
""" set hyperparameters from given covariance """
return self.covff.setCovariance(cov)
#####################
# Cached
#####################
@cached('covar_base')
def K(self):
C = self.covff.K()
RV = C[0, 0] * self._K
RV += C[0, 1] * (self._KZ + self._ZK)
RV += C[1, 1] * self._ZKZ
return RV
@cached('covar_base')
def K_grad_i(self, i):
Cgrad = self.covff.K_grad_i(i)
RV = Cgrad[0, 0] * self._K
RV += Cgrad[0, 1] * (self._KZ + self._ZK)
RV += Cgrad[1, 1] * self._ZKZ
return RV
####################
# Interpretable Params
####################
def getInterParams(self):
return self.covff.getInterParams()
def K_grad_interParam_i(self, i):
Cgrad = self.covff.K_grad_interParam_i(i)
RV = Cgrad[0, 0] * self._K
RV += Cgrad[0, 1] * (self._KZ + self._ZK)
RV += Cgrad[1, 1] * self._ZKZ
return RV
def setFIinv(self, value):
self.covff.setFIinv(value)
def getFIinv(self):
return self.covff.getFIinv()
class DirIndirCov(Covariance):
"""
Covariance matrix for decomposing direct and social genetic effects
"""
def __init__(self, kinship, design, kinship_cm = None, kinship_cross = None, jitter = 1e-4):
if kinship_cm is None: kinship_cm = kinship
if kinship_cross is None: kinship_cross = kinship
ff_dim = 2
self.covff = FreeFormCov(ff_dim, jitter = 1e-4)
self._K = kinship
self._ZK = sp.dot(design, kinship_cross.T)
self._KZ = sp.dot(kinship_cross, design.T)
self._ZKZ = sp.dot(design, sp.dot(kinship_cm, design.T))
Covariance.__init__(self, kinship.shape[0])
def dirIndirCov_K(self):
return self.covff.K()
def dirIndirCov_K_ste(self):
return self.covff.K_ste()
#####################
# Properties
#####################
@property
def variance(self):
return self.covff.variance
@property
def correlation(self):
return self.covff.correlation
#####################
# Params handling
#####################
def getParams(self):
return self.covff.getParams()
def setParams(self,params):
self.covff.setParams(params)
self.clear_all()
def getNumberParams(self):
return self.covff.getNumberParams()
def setCovariance(self,cov):
""" set hyperparameters from given covariance """
return self.covff.setCovariance(cov)
#####################
# Cached
#####################
@cached('covar_base')
def K(self):
C = self.covff.K()
RV = C[0,0] * self._K
RV += C[0,1] * (self._KZ + self._ZK)
RV += C[1,1] * self._ZKZ
return RV
@cached('covar_base')
def K_grad_i(self,i):
Cgrad = self.covff.K_grad_i(i)
RV = Cgrad[0,0] * self._K
RV += Cgrad[0,1] * (self._KZ + self._ZK)
RV += Cgrad[1,1] * self._ZKZ
return RV
####################
# Interpretable Params
####################
def getInterParams(self):
return self.covff.getInterParams()
def K_grad_interParam_i(self,i):
Cgrad = self.covff.K_grad_interParam_i(i)
RV = Cgrad[0,0] * self._K
RV += Cgrad[0,1] * (self._KZ + self._ZK)
RV += Cgrad[1,1] * self._ZKZ
return RV
def setFIinv(self, value):
self.covff.setFIinv(value)
def getFIinv(self):
return self.covff.getFIinv()
