UN_dispersion_sd: Standard deviation of a matrix
In polgia0/Chemometricmenus: Chemometric package

Description Usage Arguments Details Value Author(s) Examples

Standard deviation of a matrix by rows or columns

1	UN_dispersion_sd(previous.name = "")

previous.name

name of the default name in the menu

Allcalculation made on the whole matrix
Row wisecalculation made by rows
Column wisecalculation made by columns

The most used measure for the spread of data is the Standard Deviation(sd(x) in R notation) which should not be mixed up with the theoretical standard deviation sigma of the normal distribution. The standard deviation is very sensitive to outliers; if the distribution of the values is skewed, not only the mean will be biased, but also the standard deviation of the sample, s, will be even more biased since the standard deviation, by definition, is the sum of the squared deviations from the mean. The standard deviation is expressed using the same units of the data. In the case of normal or approximately normal distributions, s is the best measure of spread because it is the most precise estimator for sigma, the population standard deviation; unfortunately in practice the standard deviation is often uncritically used instead of taking into consideration other more robust measures for the spread of the sample values (see below Robust estimators in the Dispersion section). R console syntax: x_sd <- sd(x)

The output is saved in the variable var.dispersion

`All`	single numeric value
`Row wise`	numeric vector
`Column wise`	numeric vector

Riccardo Leardi and Gianmarco Polotti with contributions from Giorgio Marubini.Gruppo di Chemiometria (Divisione di Chimica Analitica della Societa' Chimica Italiana)