tidynorm (version 0.4.0)

reframe_with_dct: Reframe with DCT

Description

Reframe data columns using the Discrete Cosine Transform

Usage

reframe_with_dct(
  .data,
  ...,
  .token_id_col = NULL,
  .by = NULL,
  .time_col = NULL,
  .order = 5
)

Value

A data frame with with the targeted DCT coefficients, along with two additional columns

.param

The nth DCT coefficient number

.n

The number of original data values

Arguments

.data

A data frame

...

<tidy-select> One or more unquoted expressions separated by commas. These should target the vowel formant.

.token_id_col

<tidy-select> The token ID column.

.by

<tidy-select> A grouping column.

.time_col

A time column.

.order

The number of DCT parameters to return. If NA, all DCT parameters will be returned.

Details

This function will tidily apply the Discrete Cosine Transform with forward normalization (see dct for more info) to the targeted columns.

Identifying tokens

The DCT only works on a by-token basis, so there must be a column that uniquely identifies (or, in combination with a .by grouping, uniquely identifies) each individual token. This column should be passed to .token_id_col.

Order

The number of DCT coefficients to return is defined by .order. The default value is 5. Larger numbers will lead to less smoothing when the Inverse DCT is applied (see idct). Smaller numbers will lead to more smoothing.

If NA is passed to .order, all DCT parameters will be returned, which when the Inverse DCT is supplied, will completely reconstruct the original data.

Sorting by Time

An optional .time_col can also be defined to ensure that the data is correctly arranged by time.

Examples

library(tidynorm)
library(dplyr)

speaker_small <- filter(
  speaker_tracks,
  id == 0
)

speaker_dct <- reframe_with_dct(
  speaker_small,
  F1:F3,
  .by = speaker,
  .token_id_col = id,
  .time_col = t
)

head(
  speaker_dct
)