Learn R Programming

JADE (version 2.0-4)

CPPdata: Cocktail Party Problem Data

Description

This data set is a toy example for the so called cocktail party problem. In this case three sounds are mixed together with one noise source using four microphones.

Usage

data("CPPdata")

Arguments

Format

A data frame with 50000 observations on the following 4 variables.

Mic1

the mixture recorded by the first microphone.

Mic2

the mixture recorded by the second microphone.

Mic3

the mixture recorded by the third microphone.

Mic4

the mixture recorded by the fourth microphone.

Details

The three original source files were kindly provided by Ella Bingham and are also available online at the following locations: https://research.ics.aalto.fi/ica/cocktail/source5.wav, https://research.ics.aalto.fi/ica/cocktail/source7.wav and https://research.ics.aalto.fi/ica/cocktail/source9.wav.

Note that the original sound files are included in the package's subfolder datafiles. In the example section we illustrate how the CPPdata was created. An example analysis of the data is given in Miettinen et al. (2017).

References

Miettinen, J., Nordhausen, K. and Taskinen, S. (2017), Blind Source Separation Based on Joint Diagonalization in R: The Packages JADE and BSSasymp, Journal of Statistical Software, 76, 1--31, <doi:10.18637/jss.v076.i02>.

Examples

Run this code
if (FALSE) {
# the data was created as follows:
library("tuneR")
S1 <- readWave(system.file("datafiles/source5.wav", package = "JADE"))
S2 <- readWave(system.file("datafiles/source7.wav", package = "JADE"))
S3 <- readWave(system.file("datafiles/source9.wav", package = "JADE"))

set.seed(321)
NOISE <- noise("white", duration = 50000)
S <- cbind(S1@left, S2@left, S3@left, NOISE@left)
S <- scale(S, center = FALSE, scale = apply(S, 2, sd))
St <- ts(S, start = 0, frequency = 8000)
p <- 4
A <- matrix(runif(p^2, 0, 1), p, p)
A

X <- tcrossprod(St, A)
Xt <- as.ts(X)

colnames(X) <- c("Mic1", "Mic2", "Mic3", "Mic4")
CPPdata <- as.data.frame(X)
}

Run the code above in your browser using DataLab