predict.sgp_sde: Predictions from drift/diffusion estimates

In citiususc/voila: Variational Inference for Langevin Equations

Description

Predictions from drift/diffusion estimates

Usage

## S3 method for class 'sgp_sde'
predict(object, newX, lognormal = FALSE,
  quantiles = c(0.05, 0.95), ...)

Arguments

object	An sgp_sde object representing an estimate of a drift/diffusion term.
newX	A matrix representing the new input values with which to predict
lognormal	Logical value indicating if the sgp_sde represents a gaussian process or a lognormal process. If the sgp_sde is a drift estimate it should be set to FALSE, whereas if it is a diffusion estimate it should be set to TRUE.
quantiles	Vector with the quantiles that the function should return to build the confidence intervals.
...	Additional parameters (currently unused).

Value

A sde_prediction with the mean and confidence intervals of the prediction. Use the plot function to plot the estimates.

See Also

See sde_vi for an example.

citiususc/voila documentation built on May 13, 2019, 7:30 p.m.