Quickstart¶
Each wrapped R package has its own submodule. Import the one you need and call
the wrapper class like the corresponding R function; arguments are converted to
R, and the fitted R object’s components become Python attributes (with .
replaced by _).
import numpy as np
from rbartpackages import BART3
x_train = np.random.randn(100, 5)
y_train = x_train[:, 0] + 0.1 * np.random.randn(100)
bart = BART3.gbart(x_train=x_train, y_train=y_train, ndpost=200)
y_pred = bart.predict(x_train) # shape (ndpost, n)
Argument names use Python underscores in place of R dots: pass x_train for
the R argument x.train. The same pattern works for the other wrappers, e.g.
rbartpackages.BART, rbartpackages.dbarts, rbartpackages.bartMachine, and
rbartpackages.missBART.
Data frames and other array types¶
With the matching extra installed, you can pass pandas / polars data frames
or jax arrays directly; they are converted to the appropriate R object. This
is required for rbartpackages.bartMachine, whose X argument must be an R
data frame.
R documentation¶
The original R documentation of each function is appended to the corresponding
wrapper class docstring, so help(BART3.gbart) shows the upstream
reference.