R/predict_vecchia.R

In deepgp: Deep Gaussian Processes using MCMC

# Function Contents -----------------------------------------------------------
# External: (see documentation in predict.R file)
#   predict.gpvec
#   predict.dgp2vec
#   predict.dgp3vec

# Predict One Layer Vecchia ---------------------------------------------------
#' @rdname predict
#' @export

predict.gpvec <- function(object, x_new, m = object$m, lite = TRUE,
                          cores = 1, ...) {
  
  tic <- proc.time()[3]
  object <- clean_prediction(object) # remove previous predictions if present
  sep <- (is.matrix(object$theta))
  if (is.numeric(x_new)) x_new <- as.matrix(x_new)
  object$x_new <- x_new
  n_new <- nrow(object$x_new)
  
  # Pre-calculate nearest neighbors for x 
  if (lite) {
    NN_x_new <- FNN::get.knnx(object$x, x_new, m)$nn.index
    x_approx <-  NULL
  } else { 
    NN_x_new <- NULL
    x_approx <- add_pred_to_approx(object$x_approx, x_new, m)
  }
  
  # Prepare clusters
  iters <- 1:object$nmcmc
  if (cores == 1) {
    chunks <- list(iters)
  } else chunks <- split(iters, sort(cut(iters, cores, labels = FALSE)))
  if (cores > detectCores()) warning("cores is greater than available nodes")
  
  cl <- makeCluster(cores)
  registerDoParallel(cl)
  
  thread <- NULL
  result <- foreach(thread = 1:cores) %dopar% {
    out <- list()
    out$mu_t <- matrix(nrow = n_new, ncol = length(chunks[[thread]]))
    if (lite) {
      out$s2_sum <- rep(0, times = n_new)
    } else out$sigma_sum <- matrix(0, nrow = n_new, ncol = n_new)
  
    # calculate predictions for each candidate MCMC iterations
    j <- 1
    for (t in chunks[[thread]]) {
      if (sep) theta <- object$theta[t, ] else theta <- object$theta[t]
      k <- krig_vec(object$y, theta, object$g[t], object$tau2[t],
                    s2 = lite, sigma = !lite, v = object$v, m = m,
                    x = object$x, x_new = x_new, NNarray_pred = NN_x_new,
                    approx = x_approx, sep = sep)
      out$mu_t[, j] <- k$mean
      if (lite) {
        out$s2_sum <- out$s2_sum + k$s2
      } else out$sigma_sum <- out$sigma_sum + k$sigma
      j <- j + 1
    } # end of t for loop
    return(out)
  } # end of foreach loop
  
  stopCluster(cl)
    
  # Group elements out of the list
  mu_t <- do.call(cbind, lapply(result, with, eval(parse(text = "mu_t"))))
  if (lite) {
    s2_sum <- Reduce("+", lapply(result, with, eval(parse(text = "s2_sum"))))
  } else {
    sigma_sum <- Reduce("+", lapply(result, with, eval(parse(text = "sigma_sum"))))
  }

  # Add variables to the output list
  mu_cov <- cov(t(mu_t))
  object$mean <- rowMeans(mu_t)
  if (lite) { 
    object$s2 <- s2_sum / object$nmcmc + diag(mu_cov)
    object$s2_smooth <- object$s2 - mean(object$g * object$tau2)
  } else {
    object$Sigma <- sigma_sum / object$nmcmc + mu_cov
    object$Sigma_smooth <- object$Sigma - diag(mean(object$g * object$tau2), n_new)
  }
  toc <- proc.time()[3]
  object$time <- object$time + (toc - tic)
  
  return(object)
}

# Predict Two Layer Vecchia ---------------------------------------------------
#' @rdname predict
#' @export
#' 
predict.dgp2vec <- function(object, x_new, m = object$m, lite = TRUE, 
                            store_latent = FALSE, mean_map = TRUE, 
                            cores = 1, ...) {
  
  tic <- proc.time()[3]
  object <- clean_prediction(object) # remove previous predictions if present
  if (is.numeric(x_new)) x_new <- as.matrix(x_new)
  object$x_new <- x_new
  n_new <- nrow(object$x_new)
  D <- ncol(object$w[[1]])
  if (!mean_map) stop("mean_map = FALSE has not yet been implemented")

  # Pre-calculate nearest neighbors for x 
  if (lite) {
    NN_x_new <- FNN::get.knnx(object$x, x_new, m)$nn.index
    x_approx <-  NULL
  } else { 
    NN_x_new <- NULL
    x_approx <- add_pred_to_approx(object$x_approx, x_new, m)
  }
  
  # Prepare clusters
  iters <- 1:object$nmcmc
  if (cores == 1) {
    chunks <- list(iters)
  } else chunks <- split(iters, sort(cut(iters, cores, labels = FALSE)))
  if (cores > detectCores()) warning('cores is greater than available nodes')
  
  cl <- makeCluster(cores)
  registerDoParallel(cl)
  
  thread <- NULL
  result <- foreach(thread = 1:cores) %dopar% {
    out <- list()
    if (store_latent) out$w_new <- list()
    out$mu_t <- matrix(nrow = n_new, ncol = length(chunks[[thread]]))
    if (lite) {
      out$s2_sum <- rep(0, times = n_new)
    } else out$sigma_sum <- matrix(0, nrow = n_new, ncol = n_new)

    j <- 1
    for (t in chunks[[thread]]) {
      w_t <- object$w[[t]]
      
      # Map x_new to w_new (separately for each dimension)
      w_new <- matrix(nrow = n_new, ncol = D)
      for (i in 1:D) {
          k <- krig_vec(w_t[, i], object$theta_w[t, i], g = eps, tau2 = 1,
                        v = object$v, m = m, x = object$x, x_new = x_new,
                        NNarray_pred = NN_x_new)
          w_new[, i] <- k$mean
      } # end of i for loop
      if (store_latent) out$w_new[[j]] <- w_new
      
      if (lite) {
        w_approx <- NULL
      } else {
        w_approx <- update_obs_in_approx(object$w_approx, w_t)
        w_approx <- add_pred_to_approx(w_approx, w_new, m)
      }

      # Map w_new to y
      k <- krig_vec(object$y, object$theta_y[t], object$g[t], object$tau2[t],
                    s2 = lite, sigma = !lite, v = object$v, m = m,
                    x = w_t, x_new = w_new, approx = w_approx)
      out$mu_t[, j] <- k$mean
      if (lite) {
        out$s2_sum <- out$s2_sum + k$s2
      } else out$sigma_sum <- out$sigma_sum + k$sigma
      j <- j + 1
    } # end of t for loop
    return(out)
  } # end of foreach statement
    
  stopCluster(cl)
  
  # Group elements out of the list
  mu_t <- do.call(cbind, lapply(result, with, eval(parse(text = "mu_t"))))
  if (lite) {
    s2_sum <- Reduce("+", lapply(result, with, eval(parse(text = "s2_sum"))))
  } else {
    sigma_sum <- Reduce("+", lapply(result, with, eval(parse(text = "sigma_sum"))))
  }
  if (store_latent) w_new <- unlist(lapply(result, with, eval(parse(text = "w_new"))), 
                                    recursive = FALSE) 
  
  # Add variables to the output list
  mu_cov <- cov(t(mu_t))
  object$mean <- rowMeans(mu_t)
  if (store_latent) object$w_new <- w_new
  if (lite) {
    object$s2 <- s2_sum / object$nmcmc + diag(mu_cov)
    object$s2_smooth <- object$s2 - mean(object$g * object$tau2)
  } else {
    object$Sigma <- sigma_sum / object$nmcmc + mu_cov
    object$Sigma_smooth <- object$Sigma - diag(mean(object$g * object$tau2), n_new)
  }
  toc <- proc.time()[3]
  object$time <- object$time + (toc - tic)
  
  return(object)
}

# Predict Three Layer Vecchia -------------------------------------------------
#' @rdname predict
#' @export

predict.dgp3vec <- function(object, x_new, m = object$m, lite = TRUE, 
                            store_latent = FALSE, mean_map = TRUE,
                            cores = 1, ...) {
  
  tic <- proc.time()[3]
  object <- clean_prediction(object) # remove previous predictions if present
  if (is.numeric(x_new)) x_new <- as.matrix(x_new)
  object$x_new <- x_new
  n_new <- nrow(object$x_new)
  D <- ncol(object$z[[1]])
  if (!mean_map) stop("mean_map = FALSE has not yet been implemented")
  
  # Pre-calculate nearest neighbors for x 
  if (lite) {
    NN_x_new <- FNN::get.knnx(object$x, x_new, m)$nn.index
    x_approx <-  NULL
  } else { 
    NN_x_new <- NULL
    x_approx <- add_pred_to_approx(object$x_approx, x_new, m)
  }
  
  # Prepare clusters
  iters <- 1:object$nmcmc
  if (cores == 1) {
    chunks <- list(iters)
  } else chunks <- split(iters, sort(cut(iters, cores, labels = FALSE)))
  if (cores > detectCores()) warning('cores is greater than available nodes')
  
  cl <- makeCluster(cores)
  registerDoParallel(cl)
  
  thread <- NULL
  result <- foreach(thread = 1:cores) %dopar% {
    out <- list()
    if (store_latent) {
      out$z_new <- list()
      out$w_new <- list()
    }
    out$mu_t <- matrix(nrow = n_new, ncol = length(chunks[[thread]]))
    if (lite) {
      out$s2_sum <- vector(length = n_new)
    } else out$sigma_sum <- matrix(0, nrow = n_new, ncol = n_new)
  
    j <- 1
    for (t in chunks[[thread]]) {
      z_t <- object$z[[t]]
      w_t <- object$w[[t]]
    
      # Map x_new to z_new (separately for each dimension)
      z_new <- matrix(nrow = n_new, ncol = D)
      for (i in 1:D) { # mean_map = TRUE only
        k <- krig_vec(z_t[, i], object$theta_z[t, i], g = eps, tau2 = 1,
                      v = object$v, m = m, x = object$x, x_new = x_new,
                      NNarray_pred = NN_x_new)
        z_new[, i] <- k$mean
      } # end of i for loop
      if (store_latent) out$z_new[[j]] <- z_new
    
      # Map z_new to w_new (separately for each dimension)
      w_new <- matrix(nrow = n_new, ncol = D)
      for (i in 1:D) {
        k <- krig_vec(w_t[, i], object$theta_w[t, i], g = eps, tau2 = 1,
                      v = object$v, m = m, x = z_t, x_new = z_new)
        w_new[, i] <- k$mean
      } # end of i for loop
      if (store_latent) out$w_new[[j]] <- w_new
    
      if (lite) {
        w_approx <- NULL
      } else {
        w_approx <- update_obs_in_approx(object$w_approx, w_t)
        w_approx <- add_pred_to_approx(w_approx, w_new, m)
      }

      # Map w_new to y
      k <- krig_vec(object$y, object$theta_y[t], object$g[t], object$tau2[t],
                    s2 = lite, sigma = !lite, v = object$v, m = m,
                    x = w_t, x_new = w_new, approx = w_approx)
      out$mu_t[, j] <- k$mean
      if (lite) {
        out$s2_sum <- out$s2_sum + k$s2
      } else out$sigma_sum <- out$sigma_sum + k$sigma
      j <- j + 1
    } # end of t for loop
    return(out)
  } # end of foreach loop
  
  stopCluster(cl)
  
  # Group elements out of the list
  mu_t <- do.call(cbind, lapply(result, with, eval(parse(text = "mu_t"))))
  if (lite) {
    s2_sum <- Reduce("+", lapply(result, with, eval(parse(text = "s2_sum"))))
  } else {
    sigma_sum <- Reduce("+", lapply(result, with, eval(parse(text = "sigma_sum"))))
  }
  if (store_latent) {
    z_new <- unlist(lapply(result, with, eval(parse(text = "z_new"))), recursive = FALSE)
    w_new <- unlist(lapply(result, with, eval(parse(text = "w_new"))), recursive = FALSE)
  }
  
  # Add variables to the output list
  mu_cov <- cov(t(mu_t))
  object$mean <- rowMeans(mu_t)
  if (store_latent) {
    object$z_new <- z_new
    object$w_new <- w_new
  }
  if (lite) {
    object$s2 <- s2_sum / object$nmcmc + diag(mu_cov)
    object$s2_smooth <- object$s2 - mean(object$g * object$tau2)
  } else {
    object$Sigma <- sigma_sum / object$nmcmc + mu_cov
    object$Sigma_smooth <- object$Sigma - diag(mean(object$g * object$tau2), n_new)
  }
  toc <- proc.time()[3]
  object$time <- object$time + (toc - tic)
  
  return(object)
}