partial_fit function

Partial fit stochastic model to new data

Runs one iteration of the stochastic optimizer on the new data passed here.

partial_fit(optimizer, X, y = NULL, weights = NULL, ...)

Arguments

• optimizer: A stochastic optimizer from this package as output by functions oLBFGS, SQN, adaQN. Will be modified in-place.
• X: Covariates to pass to the user-defined gradient / objective / Hessian-vector functions.
• y: Target variable to pass to the user-defined gradient / objective / Hessian-vector functions.
• weights: Target variable to pass to the user-defined gradient / objective / Hessian-vector functions.
• ...: Additional arguments to pass to the user-defined gradient / objective / Hessian-vector functions.

Returns

No return value (object is modified in-place).

See Also

oLBFGS , SQN , adaQN