R Graphical Manual

Browse All

Last data update: 2014.03.03

R: ACD simulation

sim_ACD

R Documentation

ACD simulation

Description

Simulates a sample from a specified ACD model and error term distribution dist. The error terms can also be sampled from residuals. The possibility of including a diurnal seasonal component in the simulated sample is included.

Usage

    
sim_ACD(N = 1000, model = "ACD", dist = "exponential", param = NULL, order = NULL,
    Nburn = 50, startX = c(1), startMu = c(1), errors = NULL, sampleErrors = TRUE, 
    roundToSec = FALSE, rm0 = FALSE, diurnalFactor = FALSE, splineObj = NULL,
    open = NULL, close = NULL)

Arguments

`N`	sample size
`model`	the class of conditional mean duration specification. One of `"ACD"`, `"LACD1"`, `"LACD2"`, `"AMACD"`,`"ABACD"`, `"SNIACD"` or `"LSNIACD"`.
`dist`	the distribution of the error terms (only if `errors` are left out). Must be one of `"exponential"`, `"weibull"`, `"burr"`, `"gengamma"` or `"genf"`.
`param`	a vector of the parameters of the DGP (data generating process).
`order`	a vector describing the order of the conditional mean duration specification, e.g. `order = c(1,1)` for an ACD(1,1) model.
`Nburn`	the number of burned observations. Used to lower the effect of the start values of the simulated series.
`startX`	a vector of values to start the simulation from.
`startMu`	a vector of conditional mean values to start the simulation from.
`errors`	a vector of error terms. If provided and `sampleErrors = TRUE` the errors will be sampled from this vector (with replacement). If instead `sampleErrors = FALSE` the error terms will be matched by the `errors` vector non stochastic (must then be of the same length as `N + Nburn`)
`sampleErrors`	logical flag, see `errors` above. Default is `TRUE`.
`roundToSec`	if `TRUE` the simulated sample will be discretized with 1 second(unit) precision.
`rm0`	if `TRUE` zero durations will be removed. Will the result in a smaller sample than `N`.
`diurnalFactor`	if `TRUE` the simulated data will include a diurnal factor. The diurnal factor is from a fitted cubic spline given as argument to `splineObj`. If the argument `splineObj` is empty, a default fitted cubic spline from `transData` using aggregation over weekdays will be used.
`splineObj`	a cubic spline return by `diurnalAdj()`. Currently only works with cubic splines fitted with weekday aggregation. Also see `diurnalFactor` above.
`open`	only used if `diurnalFactor = TRUE` and a `splineObj` were provided. The time the exchange opens trading (as used in the fitted `splineObj`), for example `open = "10:00:00"`.
`close`	only used if `diurnalFactor = TRUE` and a `splineObj` were provided. The time the exchange close trading (as used in the fitted `splineObj`), for example `close = "18:25:00"`.

Value

a numerical vector of simulated ACD durations

Author(s)

Markus Belfrage

Examples


x <- sim_ACD() #simulates 1000 observations from an ACD(1,1) with exp. errors as default
acdFit(x)

Results


R version 3.3.1 (2016-06-21) -- "Bug in Your Hair"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> library(ACDm)
> png(filename="/home/ddbj/snapshot/RGM3/R_CC/result/ACDm/sim_ACD.Rd_%03d_medium.png", width=480, height=480)
> ### Name: sim_ACD
> ### Title: ACD simulation
> ### Aliases: sim_ACD
> 
> ### ** Examples
> 
> 
> x <- sim_ACD() #simulates 1000 observations from an ACD(1,1) with exp. errors as default
> acdFit(x)

ACD model estimation by (Quasi) Maximum Likelihood 

Call:
  acdFit(durations = x) 

Model:
  ACD(1, 1)

Distribution:
  exponential

N: 1000

Parameter estimate:
        Coef     SE    PV robustSE
omega  0.135 0.0434 0.002   0.0420
alpha1 0.147 0.0271 0.000   0.0262
beta1  0.777 0.0435 0.000   0.0414


The fixed/unfree mean distribution parameter: 
 lambda: 1

QML robust correlations:
        omega alpha1  beta1
omega   1.000  0.464 -0.862
alpha1  0.464  1.000 -0.826
beta1  -0.862 -0.826  1.000


Goodness of fit:
                     value
LogLikelihood -1505.896464
AIC            3017.792929
BIC            3032.516195
MSE               3.401044

Convergence: 0 

Number of log-likelihood function evaluations: 88 

Estimation time: 0.0177 secs 

Description: Estimated at 2016-07-04 14:06:34 by user ddbj

> 
> 
> 
> 
> 
> 
> 
> dev.off()
null device 
          1 
>