AnalyzeTS (version 2.0)

av.res: Average residuals

Description

This function calculate to return answer which are 7 accuracy of forecasting models. They are ME, MAE, MPE, MAPE, MSE, RMSE and U.

Usage

av.res(Y = NULL, F = NULL, E = NULL, r = 3)

Arguments

Y

Observation series.

F

Forecasting series.

E

Residual series.

r

Rounds the answer to the specified number of decimal places (default 3).

Value

ME

Mean Error

MAE

Mean Absolute Error

MPE

Mean Percent Error (unit: %)

MAPE

Mean Absolute Percent Error (unit: %)

MSE

Mean Square Error

RMSE

Root of Mean Square Error

U

Number Theil U

Details

The Yt is 'observation series'. The Ft is 'Forecasting series'. The et is 'residual series'. The n is size of sample. The accuracies are calculated by theory:

ME = sum(et)/n

MAE = sum(|et|)/n

MPE = sum((et/Yt)*100)/n

MAPE = sum((|et|/Yt)*100)/n

MSE = sum(et*et)/n

RMSE = sqrt(sum(et*et)/n)

U = sqrt(sum((Yt-Ft)*(Yt-Ft)))/sqrt((Yt-Y(t-1))*(Yt-Y(t-1)))

References

http://www.tailieu.tv/tai-lieu/bai-giang-quy-trinh-du-bao-khao-sat-du-lieu-va-lua-chon-mo-hinh-22414/

Examples

#Calculate moving average
library(TTR)
data(ttrc)
sma.200<-as.ts(SMA(ttrc[,"Close"],200))
ema.200<-as.ts(EMA(ttrc[,"Close"],200))
dema.200<-as.ts(DEMA(ttrc[,"Close"],200))
wma.200<-as.ts(WMA(ttrc[,"Close"],200))
evwma.200<-as.ts(EVWMA(ttrc[,"Close"],ttrc[,"Volume"],200))
zlema.200<-as.ts(ZLEMA(ttrc[,"Close"],200))
vwap.200<-as.ts(VWAP(ttrc[,"Close"],ttrc[,"Volume"],200))

#Translate series to data frame
chuoigoc<-data.frame(ttrc[,"Close"])
dubao<-data.frame(sma.200,ema.200,dema.200,wma.200,evwma.200,zlema.200,vwap.200)

#Comparing forecasting models
av.res(Y=chuoigoc,F=dubao,r=5)