I am new to pyro. I have a question regarding having multiple observations of the same GP (in this case with a linear mean function). An example of such a process is simulated below (used pymc3 as I am not sure how to evaluate a cov function with pyro).
import pymc3 as pm import numpy as np import matplotlib.pyplot as plt lengthscale = 0.2 eta = 2.0 cov = eta ** 2 * pm.gp.cov.ExpQuad(1, lengthscale) X = np.arange(10)[:, None] K = cov(X).eval() gpl = pm.MvNormal.dist(mu=np.zeros(K.shape), cov=K).random(size=10).T # linear model slope = 1.2 intercept = 2 x_train = np.random.normal(pm.MvNormal.dist(mu=np.squeeze(X)*slope + intercept, cov=K).random(size=4).T, 0.1) plt.figure(figsize=(14, 4)) plt.plot(X, x_train) plt.title("time series generated") plt.ylabel("y") plt.xlabel("X");
I have been reading through the documentation and I did not find a way to deal with multiple observations of the same GP. Is it possible in pyro?