BNN: Create the actual BNN model

In enweg/quickbnnr: Quick way to implement Bayesian Neural Networks in R and estimating these using Julia's Turing

Create the actual BNN model

Description

This function follows after make_net and supplies the BNN with the data and precompiles it

Usage

BNN(
  net,
  y,
  x,
  likelihood = c("normal", "normal_seq_to_one", "tdist", "tdist_seq_to_one"),
  ...
)

Arguments

net	A BNN network created usign make_net
y	a numerical vector of length N
x	a numerical matrix with each column being one input and one row being one variable For example, for a network with one input, N = 100 observations, nrow(x) = 1, ncol(x) = N
likelihood	determine the likelihood term of the model. See QuickBNN.jl for implemented likelihoods

enweg/quickbnnr documentation built on April 15, 2022, 3:29 a.m.