moments: Function to calculate statistical moments

In emuR: Main Package of the EMU Speech Database Management System

Function to calculate statistical moments

Description

The function calculates the first 4 moments, i.e. the mean, variance, skew, kurtosis.

Usage

moments(count, x, minval = FALSE)

Arguments

count	A vector of the observed instances per class
x	A vector of the same length as count defining the class. If missing, and if count is of class spectral, then x is equal to trackfreq(count). If x is missing and is not of class spectral, then x default to 0:(length(count)-1)
minval	If TRUE, subtract min(count) from count so that the minimum value of count is zero. This is principally used in calculating spectral moments where count is in decibels, and more generally if count contains negative values.

Details

The units of the first moment are the same as x, the units of the second moment are x\mbox{\textasciicircum}2, and the third and fourth moments are dimensionless.

Author(s)

Jonathan Harrington

References

Snedecor, G & Cochran, W. 'Statistical Methods' Iowa State Press. Wuensch,K., 2005

Examples

# first four moments of a vector
mom <- moments(bridge[,2])
# the above is the same as moments(bridge[,2], 0:12)
# first four moments of a spectral vector with the dB values
# reset so that the minimum dB value is 0. The d.c. offset is also
# excluded in the calculation
mom <- moments(e.dft[-1], minval=TRUE)
# the temporal skew of F1 for the 10th segment. Use
m <- moments(vowlax.fdat[10,1]$data)[3]

emuR documentation built on Nov. 4, 2023, 1:06 a.m.