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

xt

Browse source
This commit is contained in:
mu 2014-02-04 10:48:27 +00:00
parent b980734fb8
commit 267e0b4f1b
1 changed files with 21 additions and 12 deletions
Download patch file Download diff file Expand all files Collapse all files

33
GPy/models/state_space_xt_sep.py
View file

@ -1,4 +1,4 @@
# Copyright (c) 2013, Mu Niu, Arno Solin.
# Copyright (c) 2013, Arno Solin, Mu Niu, Simo Sarkka.
# Licensed under the BSD 3-clause license (see LICENSE.txt) ??
#
# This implementation of converting GPs to state space models is based on the article:
@ -12,7 +12,16 @@
# number = {4},
# pages = {51--61}
# }
#
# Input parameter :
# X: temporal coordinate of data point Y: spatio-temporal data SXP: spatial coordinate grid
# SI: indicate the spatial coordinate of the data point from the spatial grid.
#
# The spatial coordinate of data point do not change over time
# Kernel structure: separatable kernel
#
# Spatial kernel : Matern32
# Temporal kernel : state space of of rbf
import numpy as np
from scipy import linalg
@ -158,7 +167,7 @@ class StateSpace_1(Model):
count = count+1
(M, P) = self.kalman_filter(F1,L1,Qc1,H1,self.sigma2,Pinf1,NX.T,NY)
#stop
stop
# Run the Rauch-Tung-Striebel smoother
#if not filter:
#(M, P) = self.rts_smoother(F,L,Qc,X.T,M,P)
@ -201,7 +210,7 @@ class StateSpace_1(Model):
# Add the noise variance to the state variance
V += self.sigma2*np.eye(m.shape[0])
stop
#stop
# Lower and upper bounds
lower = m - 2*np.sqrt(V)
upper = m + 2*np.sqrt(V)
@ -244,17 +253,17 @@ class StateSpace_1(Model):
pb.imshow(Y,interpolation="nearest")
#realisation
for i in range(0,Y.shape[1]):
reli[:,i] = np.random.multivariate_normal(m[:,i],v[:,:,i])
pb.figure(3)
pb.imshow(reli,interpolation="nearest")
#for i in range(0,Y.shape[1]):
# reli[:,i] = np.random.multivariate_normal(m[:,i],v[:,:,i])
#pb.figure(3)
#pb.imshow(reli,interpolation="nearest")
for i in range(0,Y.shape[1]):
reli[:,i] = np.random.multivariate_normal(m[:,i],v[:,:,i])
pb.figure(4)
pb.imshow(reli,interpolation="nearest")
#for i in range(0,Y.shape[1]):
# reli[:,i] = np.random.multivariate_normal(m[:,i],v[:,:,i])
#pb.figure(4)
#pb.imshow(reli,interpolation="nearest")
stop
#lower = m - 2*np.sqrt(v)
#upper = m + 2*np.sqrt(v)