vv {decisionSupport} | R Documentation |
value varier function
Description
Many variables vary over time and it may not be desirable to ignore this variation in time series analyses. This function produces time series that contain variation from a specified mean and a desired coefficient of variation. A trend can be added to this time series
Usage
vv(
var_mean,
var_CV,
n,
distribution = "normal",
absolute_trend = NA,
relative_trend = NA,
lower_limit = NA,
upper_limit = NA
)
Arguments
var_mean |
mean of the variable to be varied |
var_CV |
desired coefficient of variation (in percent) |
n |
integer; number of values to produce |
distribution |
probability distribution for the introducing variation. Currently only implemented for "normal" |
absolute_trend |
absolute increment in the var_mean in each time step. Defaults to NA, which means no such absolute value trend is present. If both absolute and relative trends are specified, only original means are used |
relative_trend |
relative trend in the var_mean in each time step (in percent). Defaults to NA, which means no such relative value trend is present. If both absolute and relative trends are specified, only original means are used |
lower_limit |
lowest possible value for elements of the resulting vector |
upper_limit |
upper possible value for elements of the resulting vector |
Details
Note that only one type of trend can be specified. If neither of the trend parameters are NA, the function uses only the original means
Value
vector of n numeric values, representing a variable time series, which initially has the mean var_mean, and then increases according to the specified trends. Variation is determined by the given coefficient of variation var_CV
Author(s)
Eike Luedeling
Examples
valvar<-vv(100,10,30)
plot(valvar)
valvar<-vv(100,10,30,absolute_trend=5)
plot(valvar)
valvar<-vv(100,10,30,relative_trend=5)
plot(valvar)