predict_within_cv {plmmr} | R Documentation |
Predict method to use in cross-validation (within cvf
)
Description
Predict method to use in cross-validation (within cvf
)
Usage
predict_within_cv(
fit,
trainX,
trainY = NULL,
testX,
og_scale_beta,
std_X_details,
type,
fbm = FALSE,
idx = 1:length(fit$lambda),
Sigma_11 = NULL,
Sigma_21 = NULL,
...
)
Arguments
fit |
A list with the components returned by |
trainX |
The training data, pre-standardization and pre-rotation |
trainY |
The training outcome, not centered. Only needed if |
testX |
A design matrix used for computing predicted values (i.e, the test data). |
og_scale_beta |
testX is on the scale of the original data, so we need the beta_vals that are untransformed to match that scale.
See |
std_X_details |
A list with 3 items:
|
type |
A character argument indicating what type of prediction should be returned. Passed from |
fbm |
Logical: is trainX an FBM object? If so, this function expects that testX is also an FBM. The two X matrices must be stored the same way. |
idx |
Vector of indices of the penalty parameter |
Sigma_11 |
Variance-covariance matrix of the training data. Extracted from |
Sigma_21 |
Covariance matrix between the training and the testing data. Extracted from |
... |
Additional optional arguments |
Details
Define beta-hat as the coefficients estimated at the value of lambda that minimizes cross-validation error (CVE). Then options for type
are as follows:
'lp' (default): uses the linear predictor (i.e., product of test data and estimated coefficients) to predict test values of the outcome. Note that this approach does not incorporate the correlation structure of the data.
'blup' (acronym for Best Linear Unbiased Predictor): adds to the 'lp' a value that represents the estimated random effect. This addition is a way of incorporating the estimated correlation structure of data into our prediction of the outcome.
Note: the main difference between this function and the predict.plmm()
method is that
here in CV, predictions are made on the standardized scale (i.e., both
the trainX and testX data come from std_X). The predict.plmm()
method
makes predictions on the scale of X (the original scale)
Value
A numeric vector of predicted values