DLflow: DLflow Class

In neuroimaginador/dl4ni: Deep Learning Flows for Neuroimaging

Description

DLflow Class

Usage

DLflow

Format

R6Class object.

Value

Object of R6Class and DLflow.

Methods

Documentation

For full documentation of each method follow the corresponding link.

initialize(name, inputs = list())

Create a new flow with the given name and inputs. Documented in DLflow.initialize.

name()

Returns the name of the flow. Documented in DLflow.name.

get_inputs()

Returns the inputs of the flow. Documented in DLflow.get_inputs.

get_outputs()

Returns the outputs of the flow. Documented in DLflow.get_outputs.

get_model(output)

Returns the DLmodel that computes the given output. Documented in DLflow.get_model.

replace(output, with)

Replace a function, scheme or model that computes an output with another function, scheme or model. Documented in DLflow.replace.

add(what = NULL, inputs = NULL, output = NULL, subset = NULL)

Adds a function, scheme or model to the flow, that transforms the given inputs into the given output. Documented in DLflow.add.

execute(inputs = list(), desired_outputs = NULL, initialize_outputs = TRUE, mode = c("debug", "faster", "medium", "slower"))

Execute the flow to obtain some outputs, given input files. Documented in DLflow.execute.

run(...)

Just a wrapper for execute. Documented in DLflow.run.

train(output, input_filenames, output_filenames, train_split = 0.75, epochs = 10, target_windows_per_file = 1024, mode = c("debug", "faster", "medium", "slower"))

Train the model to compute an output. Documented in DLflow.train.

graph()

Returns the graph of the flow. Documented in DLflow.graph.

plot(interactive = FALSE)

Plots the graph of the flow. Documented in DLflow.plot.

reset(outputs = 'all')

Remove the specified computed outputs. Documented in DLflow.reset.

save(path = tempdir(), file_prefix = self$name())

Save the flow to disk. Documented in DLflow.save.

load(filename)

Load a flow from disk. Documented in DLflow.load.

subset(outputs)

Take a subset of the flow, for given outputs. Documented in DLflow.subset.

clone(deep = FALSE)

Clone a flow. Documented in DLflow.clone.

Examples

DLflow$new()

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