DataLoader.Seq: R6 class representign a data loader for sequence datasets.

In modello: Homemade Deep Learning Library

Description

R6 class representign a data loader for sequence datasets.

R6 class representign a data loader for sequence datasets.

Details

It loads the data from a R array

Methods

Public methods

• DataLoader.Seq$new()

• DataLoader.Seq$length()

• DataLoader.Seq$has.next()

• DataLoader.Seq$reset()

• DataLoader.Seq$feed()

• DataLoader.Seq$clone()

---

Method new()

Initialisation method

Usage

DataLoader.Seq$new(ds, bz)

Arguments

ds

sequence datast

bz

integer, batch size

---

Method length()

Returns the length of teh data loader object

Usage

DataLoader.Seq$length()

Examples

ds = DataSet.Seq$new(matrix(rnorm(100), 20), 5, 1)
dl = DataLoader.Seq$new(ds, 5)
ds$length()
dl$length()

---

Method has.next()

Returns TRUE if the data loader has still batches available

Usage

DataLoader.Seq$has.next()

Examples

ds = DataSet.Seq$new(matrix(rnorm(100), 20), 5, 1)
dl = DataLoader.Seq$new(ds, 5)
ds$has.next() # TRUE

---

Method reset()

Resets the data loader to its initial state

Usage

DataLoader.Seq$reset()

---

Method feed()

Feeds the data into a sequence of batches of sequences or it creates a new one

Usage

DataLoader.Seq$feed(SEQ, mode = "head")

Arguments

SEQ

batch of sequences. If NULL it creates a new one

mode

feeding mode

Examples

modello.init(100, 100, 100, 100)
ds = DataSet.Seq$new(matrix(rnorm(100), 20), 5, 1)
dl = DataLoader.Seq$new(ds, 5)
SEQ = dl$feed(NULL)
print(SEQ)
print(SEQ$X[[1]][[1]]$v)
dl$feed(SEQ)
print(SEQ$X[[1]][[1]]$v)
modello.close()

---

Method clone()

The objects of this class are cloneable with this method.

Usage

DataLoader.Seq$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Author(s)

Filippo Monari

Examples

## ------------------------------------------------
## Method `DataLoader.Seq$length`
## ------------------------------------------------

ds = DataSet.Seq$new(matrix(rnorm(100), 20), 5, 1)
dl = DataLoader.Seq$new(ds, 5)
ds$length()
dl$length()

## ------------------------------------------------
## Method `DataLoader.Seq$has.next`
## ------------------------------------------------

ds = DataSet.Seq$new(matrix(rnorm(100), 20), 5, 1)
dl = DataLoader.Seq$new(ds, 5)
ds$has.next() # TRUE

## ------------------------------------------------
## Method `DataLoader.Seq$feed`
## ------------------------------------------------

modello.init(100, 100, 100, 100)
ds = DataSet.Seq$new(matrix(rnorm(100), 20), 5, 1)
dl = DataLoader.Seq$new(ds, 5)
SEQ = dl$feed(NULL)
print(SEQ)
print(SEQ$X[[1]][[1]]$v)
dl$feed(SEQ)
print(SEQ$X[[1]][[1]]$v)
modello.close()

modello documentation built on Feb. 2, 2021, 9:06 a.m.