pred.compareModels: Compare different predictions
In predtoolsTS: Time Series Prediction Tools

Description Usage Arguments Details Author(s) Examples

Plots the original time serie along with 2-5 predictive models.

1
2
3

pred.compareModels(originalTS, p_1, p_2, p_3 = NULL, p_4 = NULL,
  p_5 = NULL, legendNames = NULL, colors = NULL, legend = TRUE,
  legendPosition = NULL, yAxis = "Values", title = "Predictions")

`originalTS`	A `ts` object
`p_1`	A `ts` object
`p_2`	A `ts` object
`p_3`	A `ts` object. Default is NULL.
`p_4`	A `ts` object. Default is NULL.
`p_5`	A `ts` object. Default is NULL.
`legendNames`	String vector with the names for the legend. Has to be same length as number of time series we are plotting(including the original one). Default is NULL.
`colors`	Vector with the colors. Has to be same length as number of time series we are plotting(including the original one). Default is NULL.
`legend`	A logical. Do we want a legend? Default is TRUE.
`legendPosition`	A string with the position of the legend (bottomright, topright, ...). Default is NULL.
`yAxis`	A string. Name for the y axis. "Values" as default.
`title`	A string. Title for the plot. Default is "Predictions".

This function aims to ease the comparation between different predictive models by plotting them into the same graphic.

Alberto Vico Moreno

data(AirPassengers)
#pre-processing
p <- prep(AirPassengers)
#modelling
arima.modl <- modl(p)
cart.modl <- modl(p,method='dataMining',algorithm='rpart')
#predicting
arima.pred <- pred(arima.modl,n.ahead=30)
cart.pred <- pred(cart.modl,n.ahead=45)
#post-processing
arima.pred <- postp(arima.pred,p)
cart.pred <- postp(cart.pred,p)
#visual comparison
pred.compareModels(AirPassengers,arima.pred$predictions,cart.pred$predictions
,legendNames=c('AirPassengers','ARIMA','CART'),yAxis='Passengers',legendPosition = 'topleft')