Time Series Analysis

Mid Exam - Week 8

Email           :
RPubs          : https://rpubs.com/invokerarts/
Github         : https://github.com/invokerarts
Majors         : Business Statistics
Address      : ARA Center, Matana University Tower Jl. CBD Barat Kav, RT.1, Curug Sangereng,
                      Kelapa Dua, Tangerang, Banten 15810.


1 Manupulasi Data

# install.packages("readxl")
# install.packages("dplyr")
# install.packages("DT")

library("dplyr")
library("readxl")
library("DT")

data <- read_xlsx("JPYtoIDR.xlsx")
datatable(data)
JPYtoIDR <- data.frame("Tanggal" = data$Tanggal, "Kurs" = data$Terakhir)
JPYtoIDR$Tanggal <- format(as.Date(JPYtoIDR$Tanggal, '%Y-%m-%d'), '%m-%Y')
JPYtoIDR$Kurs    <- sub(",", ".", JPYtoIDR$Kurs) %>% as.numeric()
datatable(JPYtoIDR)

2 X-12-ARIMA & STL

2.1 Membuat Time Series

library("x12")
DataJPYIDR <-ts(rev(JPYtoIDR$Kurs),start=c(2015,10),frequency = 12) 
DataJPYIDR
##         Jan    Feb    Mar    Apr    May    Jun    Jul    Aug    Sep    Oct
## 2015                                                                113.35
## 2016 113.73 118.59 117.73 123.90 123.31 127.84 128.29 128.25 128.72 124.46
## 2017 118.30 118.17 119.59 119.46 120.24 118.54 120.83 121.30 119.72 119.31
## 2018 122.57 128.80 129.46 127.22 127.63 129.40 128.83 132.56 131.04 134.56
## 2019 128.28 126.21 128.39 127.83 131.78 130.90 128.82 133.37 131.28 129.87
## 2020 125.91 132.65 151.54 138.29 135.17 131.36 137.23 137.46 140.74 139.68
## 2021 133.89 133.57 131.13 132.11 130.28 130.44 131.78 129.62 128.04       
##         Nov    Dec
## 2015 112.33 114.55
## 2016 118.35 115.21
## 2017 120.16 120.36
## 2018 126.00 131.17
## 2019 128.74 127.76
## 2020 135.09 135.95
## 2021

2.2 Mencari lag sebelum X-12-ARIMA

par(mfrow=c(1,2)) 
acf(DataJPYIDR, lag.max = 48, main = "Sample ACF for the mean monthly DataJPYIDR series")
pacf(DataJPYIDR, lag.max = 48, main = "Sample PACF for the mean monthly DataJPYIDR series")

2.3 X-12-ARIMA

JPYIDRX12 <- x12(DataJPYIDR)
plot(JPYIDRX12,original=T,trend=T,sa=T)

2.4 STL Data

JPYIDRSTL <- stl(DataJPYIDR, t.window=15, s.window="periodic", robust=TRUE)
plot(JPYIDRSTL)

3 Forecasting

# install.packages("forecast")
library(forecast)
FORECAST <-forecast(JPYIDRSTL, method = "naive")
plot(FORECAST, ylab="New orders index")

window(FORECAST$mean,start=c(2021,09),end=c(2022,3))
##           Jan      Feb      Mar Apr May Jun Jul Aug Sep      Oct      Nov
## 2021                                                    126.7099 123.3282
## 2022 123.2181 124.9762 124.6525                                          
##           Dec
## 2021 123.9749
## 2022

4 Conclusion

Jadi, Perkiraan paling akurat untuk kurs JPY-IDR bulan Oktober, November, Desember, Januari, Februari, Maret tahun 2021-2022, yaitu: 126.7099, 123.3282, 123.9749, 123.2181, 124.9762, dan 124.6525 masing-masing.