modelOutputGrad: Compute derivatives with respect to params of model outputs.
Description
Compute derivatives with respect to params of model outputs.
Usage
g <- modelOutputGrad(model, X)
g <- modelOutputGrad(model, X, dim)
Arguments
model
the model structure for which gradients are computed.
X
input locations where gradients are to be computed.
dim
the dimension of the model for which gradients are required.
Value
ggradients of the model output with respect to the model parameters for the given input locations. The size of the returned matrix is of dimension number of data x number of parameters x number of model outputs (which maintains compatability with NETLAB).
Details
g <- modelOutputGrad(model, X)
gives the gradients of the outputs from the model with respect to the parameters for a given set of inputs.
g <- modelOutputGrad(model, X, dim)
gives the gradients of the outputs from the model with respect to the parameters for a given set of inputs.