tfdatasets (version 2.2.0)

dataset_window: Combines input elements into a dataset of windows.

Description

Combines input elements into a dataset of windows.

Usage

dataset_window(dataset, size, shift = NULL, stride = 1, drop_remainder = FALSE)

Arguments

dataset

A dataset

size

representing the number of elements of the input dataset to combine into a window.

shift

epresenting the forward shift of the sliding window in each iteration. Defaults to size.

stride

representing the stride of the input elements in the sliding window.

drop_remainder

representing whether a window should be dropped in case its size is smaller than window_size.

See Also

Other dataset methods: dataset_batch(), dataset_cache(), dataset_collect(), dataset_concatenate(), dataset_decode_delim(), dataset_filter(), dataset_interleave(), dataset_map_and_batch(), dataset_map(), dataset_padded_batch(), dataset_prefetch_to_device(), dataset_prefetch(), dataset_reduce(), dataset_repeat(), dataset_shuffle_and_repeat(), dataset_shuffle(), dataset_skip(), dataset_take()