Predictions from serialized dbarts model fits

[jeremyrcoyle]

I'm trying to save and reload dbarts fits. I saw the section in the help explaining the necessary step of touching the state object:

Saving
saveing and loading fitted BART objects for use with predict requires that R's serialization mechanism be able to access the underlying trees, in addition to being fit with keeptrees/keepTrees as TRUE. For memory purposes, the trees are not stored as R objects unless specifically requested. To do this, one must “touch” the sampler's state object before saving, e.g. for a fitted object bartFit, execute invisible(bartFit$fit$state).

However, saving and loading still breaks predictions even with this step. Here's an example:

# example from dbarts help
library(dbarts)
## simulate data (example from Friedman MARS paper)
## y = f(x) + epsilon , epsilon ~ N(0, sigma)
## x consists of 10 variables, only first 5 matter

f <- function(x) {
  10 * sin(pi * x[,1] * x[,2]) + 20 * (x[,3] - 0.5)^2 +
    10 * x[,4] + 5 * x[,5]
}

set.seed(99)
sigma <- 1.0
n     <- 100

x  <- matrix(runif(n * 10), n, 10)
Ey <- f(x)
y  <- rnorm(n, Ey, sigma)

## run BART
set.seed(99)
bartFit <- bart(x, y, keeptrees=TRUE)

# generate predictions
preds <- predict(bartFit, x)


# touch the state object
invisible(bartFit$fit$state)

# attempt to save and load
tmp <- tempfile()
save(bartFit, file=tmp)
rm(bartFit)
load(tmp)

# new predictions are NaN
preds_after_save <- predict(bartFit, x)
all(is.nan(preds_after_save))

Are there additional steps necessary for this to work?

[vdorie]

Should be fixed with ead2786. Thanks for letting me know!