cuda.ml (version 0.3.2)

cuda_ml_ols: Train a OLS model.

Description

Train an Ordinary Least Square (OLS) model for regression tasks.

Usage

cuda_ml_ols(x, ...)
# S3 method for default
cuda_ml_ols(x, ...)
# S3 method for data.frame
cuda_ml_ols(
  x,
  y,
  method = c("svd", "eig", "qr"),
  fit_intercept = TRUE,
  normalize_input = FALSE,
  ...
)
# S3 method for matrix
cuda_ml_ols(
  x,
  y,
  method = c("svd", "eig", "qr"),
  fit_intercept = TRUE,
  normalize_input = FALSE,
  ...
)
# S3 method for formula
cuda_ml_ols(
  formula,
  data,
  method = c("svd", "eig", "qr"),
  fit_intercept = TRUE,
  normalize_input = FALSE,
  ...
)
# S3 method for recipe
cuda_ml_ols(
  x,
  data,
  method = c("svd", "eig", "qr"),
  fit_intercept = TRUE,
  normalize_input = FALSE,
  ...
)

Arguments

x

Depending on the context:

* A __data frame__ of predictors. * A __matrix__ of predictors. * A __recipe__ specifying a set of preprocessing steps * created from [recipes::recipe()]. * A __formula__ specifying the predictors and the outcome.

...

Optional arguments; currently unused.

y

A numeric vector (for regression) or factor (for classification) of desired responses.

method

Must be one of "svd", "eig", "qr".

- "svd": compute SVD decomposition using Jacobi iterations. - "eig": use an eigendecomposition of the covariance matrix. - "qr": use the QR decomposition algorithm and solve `Rx = Q^T y`.

If the number of features is larger than the sample size, then the "svd" algorithm will be force-selected because it is the only algorithm that can support this type of scenario.

Default: "svd".

fit_intercept

If TRUE, then the model tries to correct for the global mean of the response variable. If FALSE, then the model expects data to be centered. Default: TRUE.

normalize_input

Ignored when fit_intercept is FALSE. If TRUE, then the predictors will be normalized to have a L2 norm of 1. Default: FALSE.

formula

A formula specifying the outcome terms on the left-hand side, and the predictor terms on the right-hand side.

data

When a __recipe__ or __formula__ is used, data is specified as a __data frame__ containing the predictors and (if applicable) the outcome.

Value

A OLS regressor that can be used with the 'predict' S3 generic to make predictions on new data points.

Examples

# NOT RUN {
library(cuda.ml)

model <- cuda_ml_ols(formula = mpg ~ ., data = mtcars, method = "qr")
predictions <- predict(model, mtcars[names(mtcars) != "mpg"])

# predictions will be comparable to those from a `stats::lm` model
lm_model <- stats::lm(formula = mpg ~ ., data = mtcars, method = "qr")
lm_predictions <- predict(lm_model, mtcars[names(mtcars) != "mpg"])

print(
  all.equal(
    as.numeric(lm_predictions),
    predictions$.pred,
    tolerance = 1e-3
  )
)
# }