block_paths: Create Independent Paths

In neuroimaginador/dl4ni: Deep Learning Flows for Neuroimaging

Description

This function appends a set of independent paths with the same architecture to a keras object.

Usage

block_paths(object, hidden_layers = NULL, num_paths = 2,
  finalize_layers = list(), concatenate = FALSE)

Arguments

object	(keras object) The object where to append the block.
hidden_layers	(list) List of layers, with types and corresponding parameters.
num_paths	(integer) Number of paths to generate with the same topology.
finalize_layers	(list) List of layers, with types and corresponding parameters, to be added at the end of each path. Useful when called from block_categorical, block_regression, etc.
concatenate	(logical) Default = FALSE. Do we have to concatenate all outputs?

Value

A list of keras objects (if concatenate == FALSE) or a single object (if concatenate == TRUE), representing the num_paths independent paths.

neuroimaginador/dl4ni documentation built on May 3, 2019, 5:47 p.m.