# bandSpectra: Average Spectra within Frequency Bands In gslcca: Generalized Semi-Linear Canonical Correlation Analysis

## Description

Given a matrix of spectra, where each row gives the power at each frequency, create banded spectra by averaging across columns within frequency bands.

## Usage

 `1` ```bandSpectra(spectra, breaks = NULL, labels = NULL, ...) ```

## Arguments

 `spectra` A matrix of spectra with numeric column names (as returned by `readSpectra`), or column names that can be coerced to numeric once any characters other than 0-9 and “.” are removed. `breaks` The break points between the frequency bands. `labels` Labels for the frequency bands. `...` Further arguments passed to cut.

## Value

A matrix of banded spectra.

## Author(s)

Heather Turner

`gslcca`
 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21``` ```## apply to arbitrary matrix to average over column groups M <- matrix(1:12, 3, 4) colnames(M) <- 1:4 ## aggregate first two and last two columns ## (colnames <= 2, or colnames > 2) bandSpectra(M, breaks = 2) ## apply to clonidine experiment data(clonidine) colnames(clonidine\$spectra) banded <- bandSpectra(clonidine\$spectra, breaks = c(4, 8, 10, 13, 15, 20, 28), labels = c("delta", "theta", "alpha1", "alpha2", "not named", "beta1", "beta2", "gamma")) ## run gslcca result <- gslcca(banded, "Critical Exponential", time = Time, treatment = Treatment, subject = Rat, separate = FALSE, subject.smooth = 4, data = clonidine) plot(result, "fitted") plot(result, "signatures") ```