apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-06-05 14:55:15 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

Now the model for each dt is calculated only once

Browse source
This commit is contained in:
Arno Solin 2013-11-11 23:48:38 +00:00
parent 53f0fbcbff
commit 2303132933
1 changed files with 50 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

65
GPy/models/state_space.py
View file

@ -217,11 +217,16 @@ class StateSpace(Model):
dt[:,0] = X[:,1]-X[:,0]
dt[:,1:] = np.diff(X)
# Solve the LTI SDE for these time steps
As, Qs, index = self.lti_disc(F,L,Qc,dt)
# Kalman filter
for k in range(0,Y.shape[1]):
# Form discrete-time model
(A, Q) = self.lti_disc(F,L,Qc,dt[:,k])
#(A, Q) = self.lti_disc(F,L,Qc,dt[:,k])
A = As[:,:,index[k]];
Q = Qs[:,:,index[k]];
# Prediction step
MF[:,k] = A.dot(MF[:,k-1])
@ -245,11 +250,16 @@ class StateSpace(Model):
dt[:,0] = X[:,1]-X[:,0]
dt[:,1:] = np.diff(X)
# Solve the LTI SDE for these time steps
As, Qs, index = self.lti_disc(F,L,Qc,dt)
# Sequentially smooth states starting from the end
for k in range(2,X.shape[1]+1):
# Form discrete-time model
(A, Q) = self.lti_disc(F,L,Qc,dt[:,1-k])
#(A, Q) = self.lti_disc(F,L,Qc,dt[:,1-k])
A = As[:,:,index[1-k]];
Q = Qs[:,:,index[1-k]];
# Smoothing step
LL = linalg.cho_factor(A.dot(PS[:,:,-k]).dot(A.T)+Q)
@ -273,11 +283,16 @@ class StateSpace(Model):
dt[:,0] = X[:,1]-X[:,0]
dt[:,1:] = np.diff(X)
# Solve the LTI SDE for these time steps
As, Qs, index = self.lti_disc(F,L,Qc,dt)
# Kalman filter for likelihood evaluation
for k in range(0,Y.shape[1]):
# Form discrete-time model
(A,Q) = self.lti_disc(F,L,Qc,dt[:,k])
#(A,Q) = self.lti_disc(F,L,Qc,dt[:,k])
A = As[:,:,index[k]];
Q = Qs[:,:,index[k]];
# Prediction step
m = A.dot(m)
@ -323,19 +338,39 @@ class StateSpace(Model):
# Dimensionality
n = F.shape[0]
index = 0
# The covariance matrix by matrix fraction decomposition
Phi = np.zeros((2*n,2*n))
Phi[:n,:n] = F
Phi[:n,n:] = L.dot(Qc).dot(L.T)
Phi[n:,n:] = -F.T
AB = linalg.expm(Phi*dt).dot(np.vstack((np.zeros((n,n)),np.eye(n))))
#Q = AB[:n,:].dot(linalg.inv(AB[n:,:]))
Q = linalg.solve(AB[n:,:].T,AB[:n,:].T)
# Check for numbers of time steps
if dt.flatten().shape[0]==1:
# The dynamical model
A = linalg.expm(F*dt)
# The covariance matrix by matrix fraction decomposition
Phi = np.zeros((2*n,2*n))
Phi[:n,:n] = F
Phi[:n,n:] = L.dot(Qc).dot(L.T)
Phi[n:,n:] = -F.T
AB = linalg.expm(Phi*dt).dot(np.vstack((np.zeros((n,n)),np.eye(n))))
Q = linalg.solve(AB[n:,:].T,AB[:n,:].T)
# Return
return (A, Q)
# The dynamical model
A = linalg.expm(F*dt)
# Return
return A, Q
# Optimize for cases where time steps occur repeatedly
else:
# Time discretizations
dt, _, index = np.unique(dt,True,True)
# Allocate space for A and Q
A = np.empty((n,n,dt.shape[0]))
Q = np.empty((n,n,dt.shape[0]))
# Call this function for each dt
for j in range(0,dt.shape[0]):
A[:,:,j], Q[:,:,j] = self.lti_disc(F,L,Qc,dt[j])
# Return
return A, Q, index