#Y = Y[0:10, :] numData = Y.shape[0] d = Y.shape[1] #v, X = netlab.pca(Y.transpose(), q) X = numpy.random.normal(0.0, 1e-1, (numData, 2)) Xstore = X; Xstor = numpy.empty(Xstore.shape) Ystore = Y; Y = ndlwrap.fromarray(Y) X = ndlwrap.fromarray(X) # Set up kernel function kern = nl.cmpndKern() kern1 = nl.rbfKern(X) kern2 = nl.biasKern(X) kern3 = nl.whiteKern(X) kern3.setVariance(1e-3) kern.addKern(kern1) kern.addKern(kern2) kern.addKern(kern3) noise = nl.gaussianNoise(Y) # Create a GP model. model = nl.gp(q, d, X, Y, kern, noise, nl.gp.DTCVAR, 100, 3) model.setBetaVals(math.exp(2)) #pdb.set_trace() model.setDefaultOptimiser(nl.gp.CG)
Y, lbls = datasets.lvmLoadData('oil100')
d = Y.shape[0]
numData = Y.shape[1]
v, X = netlab.pca(Y.transpose(), q)
X = numpy.random.normal(0.0, 1e-6, (Y.shape[0], 2))
Xstore = X;
Xstor = numpy.empty(Xstore.shape)
Ystore = Y;
Y = ndlwrap.fromarray(Y)
X = ndlwrap.fromarray(X)
# Set up kernel function
kern = ndlml.cmpndKern()
kern1 = ndlml.rbfKern(X)
kern2 = ndlml.biasKern(X)
kern3 = ndlml.whiteKern(X)
kern.addKern(kern1)
kern.addKern(kern2)
kern.addKern(kern3)
noise = ndlml.gaussianNoise(Y)
# Create an GP model.
model = ndlml.gp(2, 12, X, Y, kern, noise, ndlml.gp.FTC, 100, 3)
model.setOptimiseX(True)
model.setDefaultOptimiser(ndlml.gp.GD)
X, y = datasets.mapLoadData('spgp1d')
numActive = 9
Xstore = np.array(X)
ystore = np.array(y)
X = nl.matrix(X)
y = nl.matrix(y)
# Set up kernel function
kern = nl.cmpndKern()
kern1 = nl.rbfKern(X)
kern2 = nl.biasKern(X)
kern3 = nl.whiteKern(X)
kern3.setVariance(1e-4)
kern.addKern(kern1)
kern.addKern(kern2)
kern.addKern(kern3)
noise = nl.gaussianNoise(y)
# Create a GP model.
model = nl.gp(1, 1, X, y, kern, noise, nl.gp.DTC, numActive, 3)
model.setBetaVal(math.exp(2))
model.setDefaultOptimiser(nl.gp.CG)
# Optimise the GP.
# set initial values to small random numbers. X = np.random.normal(0.0, 1e-6, (len(tr.movieIDs()), q)) X_u = np.random.normal(0.0, 1e-6, (numActive, q)) lnsigma2 = math.log(startVar) lnbeta = -math.log(startVar) # Set up kernel functions kernFtc = nl.cmpndKern(q) kernSp = nl.cmpndKern(q) kern1 = nl.rbfKern(q) kern2 = nl.biasKern(q) kern3 = nl.whiteKern(q) kern4 = nl.whitefixedKern(q) kern2.setVariance(0.11) kern3.setVariance(math.exp(lnsigma2)) kern4.setVariance(1e-2) kernFtc.addKern(kern1) kernFtc.addKern(kern2) kernFtc.addKern(kern3) kernSp.addKern(kern1) kernSp.addKern(kern2) kernSp.addKern(kern4)