Scribd Logo
Upload

Welcome to Scribd!

  • 95 views

    ARIMA

    Uploaded by

    Quinn Avril
    This document reports on the analysis of time series coffee production data. The data is read in and converted to log format. An ARIMA model is fitted to the log-transformed data and used to forecast future coffee production values for the next 3 years. Diagnostic tests, including the Box-Ljung test, are performed on the residuals of the ARIMA model to check that the residuals are independently distributed.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    ARIMA

    Uploaded by

    Quinn Avril
    95 views8 pages
    This document reports on the analysis of time series coffee production data. The data is read in and converted to log format. An ARIMA model is fitted to the log-transformed data and used to forecast future coffee production values for the next 3 years. Diagnostic tests, including the Box-Ljung test, are performed on the residuals of the ARIMA model to check that the residuals are independently distributed.

    Original Title

    ARIMA.docx

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document reports on the analysis of time series coffee production data. The data is read in and converted to log format. An ARIMA model is fitted to the log-transformed data and used to forecast future coffee production values for the next 3 years. Diagnostic tests, including the Box-Ljung test, are performed on the residuals of the ARIMA model to check that the residuals are independently distributed.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    95 views8 pages

    ARIMA

    Uploaded by

    Quinn Avril
    This document reports on the analysis of time series coffee production data. The data is read in and converted to log format. An ARIMA model is fitted to the log-transformed data and used to forecast future coffee production values for the next 3 years. Diagnostic tests, including the Box-Ljung test, are performed on the residuals of the ARIMA model to check that the residuals are independently distributed.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    of 8

    > View(mydata)

    > mydata<- read.csv("file:///C:/Users/JENSENATICS/Desktop/thesis/Coffeeproduction2.csv")


    > attach(mydata)
    > View(mydata)
    > library(MASS)
    > library(tseries)
    ‘tseries’ version: 0.10-46

    ‘tseries’ is a package for time series analysis and computational finance.

    See ‘library(help="tseries")’ for details.


    > library(forecast)
    This is forecast 8.5
    Use suppressPackageStartupMessages() to eliminate package startup messages.
    > lnstock=log(C[1:11])
    Error in C[1:11] : object of type 'closure' is not subsettable
    > #plot and convert to ln format
    > lnstock=log(C[1:11])
    Error in C[1:11] : object of type 'closure' is not subsettable
    > #plot and convert to ln format
    > lnstock=log(C[1:11])
    Error in C[1:11] : object of type 'closure' is not subsettable
    > lnstock
    Error: object 'lnstock' not found
    > #plot and convert to ln format
    > lnstock=log(C[1.11])
    Error in C[1.11] : object of type 'closure' is not subsettable
    > #plot and convert to ln format
    > lnstock=log(CP[1.11])
    > #plot and convert to ln format
    > lnstock=log(CP[1:11])
    > lnstock
    [1] 11.49146 11.48687 11.47660 11.45674 11.39105 11.39575 11.27256 11.23128 11.18916
    [10] 11.13929 11.03615
    > acf(lnstock,lag.max = 11)
    > pacf(lnstock,lag.max = 11)
    > difflnstock=diff(lnstock,1)
    > adf.test(lnstock)

    Augmented Dickey-Fuller Test


    data: lnstock
    Dickey-Fuller = -1.8042, Lag order = 2, p-value = 0.647
    alternative hypothesis: stationary

    > difflnstock=diff(lnstock,2)
    > difflnstock=diff(lnstock,0)
    Error in diff.default(lnstock, 0) :
    'lag' and 'differences' must be integers >= 1
    > difflnstock=diff(lnstock,1)
    > adf.test(lnstock)

    Augmented Dickey-Fuller Test

    data: lnstock
    Dickey-Fuller = -1.8042, Lag order = 2, p-value = 0.647
    alternative hypothesis: stationary

    > adf.test(difflnstock)

    Augmented Dickey-Fuller Test


    data: difflnstock
    Dickey-Fuller = -1.7894, Lag order = 2, p-value = 0.6526
    alternative hypothesis: stationary
    >
    > pricearima<- ts(lnstock, start=c(2007,1),frequency = 11)
    > pricearima<- ts(lnstock, start=c(2007,2017),frequency = 11)
    > pricearima<- ts(lnstock, start=c(2007,20),frequency = 11)
    > pricearima<- ts(lnstock, start=c(2007,25),frequency = 11)
    > pricearima<- ts(lnstock, start=c(2007,30),frequency = 11)
    > pricearima<- ts(lnstock, start=c(2007,90),frequency = 11)
    > pricearima<- ts(lnstock, start=c(2007,30),frequency = 11)
    > pricearima<- ts(lnstock, start=c(2007,1),frequency = 11)
    > pricearima<- ts(lnstock, start=c(2007,90),frequency = 11)
    > pricearima<- ts(lnstock, start=c(2007,100),frequency = 11)
    > fitlnstock<- auto.arima(pricearima)
    Warning message:
    In value[[3L]](cond) :
    The chosen test encountered an error, so no seasonal differencing is selected. Check th
    time series data.
    > pricearima<- ts(lnstock, start=c(2007,0),frequency = 11)
    > pricearima<- ts(lnstock, start=c(2017,0),frequency = 11)
    > fitlnstock<- auto.arima(pricearima)
    Warning message:
    In value[[3L]](cond) :
    The chosen test encountered an error, so no seasonal differencing is selected. Check th
    time series data.
    > pricearima<- ts(lnstock, start=c(2007,1),frequency = 1)
    > fitlnstock<- auto.arima(pricearima)
    > fitlnstock
    Series: pricearima
    ARIMA(0,1,0) with drift

    Coefficients:
    drift
    -0.0455
    s.e. 0.0126

    sigma^2 estimated as 0.001782: log likelihood=18.03


    AIC=-32.06 AICc=-30.34 BIC=-31.45
    > plot(pricearima,type = "l")
    > title('jb c')
    > productionarima<- ts(lnstock, start=c(2007,1),frequency = 1)
    > fitlnstock<- auto.arima(productionarima)
    > fitlnstock
    Series: productionarima
    ARIMA(0,1,0) with drift

    Coefficients:
    drift
    -0.0455
    s.e. 0.0126

    sigma^2 estimated as 0.001782: log likelihood=18.03


    AIC=-32.06 AICc=-30.34 BIC=-31.45
    > plot(productionarima,type = "l")
    > title('Coffee production')
    > exp(lnstock)
    [1] 97876.68 97427.99 96432.95 94536.01 88526.09 88943.00 78633.92 75454.49 72341.82
    [10] 68822.93 62077.95
    > forecastedvalues_ln=forecast(fitlnstock,h=11)
    > forecastedvalues_ln
    Point Forecast Lo 80 Hi 80 Lo 95 Hi 95
    2018 10.99061 10.93651 11.04472 10.90787 11.07336
    2019 10.94508 10.86857 11.02159 10.82807 11.06210
    2020 10.89955 10.80584 10.99326 10.75624 11.04286
    2021 10.85402 10.74582 10.96222 10.68854 11.01950
    2022 10.80849 10.68751 10.92946 10.62347 10.99350
    2023 10.76296 10.63043 10.89548 10.56028 10.96563
    2024 10.71742 10.57428 10.86056 10.49851 10.93634
    2025 10.67189 10.51887 10.82492 10.43786 10.90592
    2026 10.62636 10.46406 10.78867 10.37814 10.87458
    2027 10.58083 10.40974 10.75191 10.31918 10.84248
    2028 10.53530 10.35586 10.71473 10.26087 10.80972
    > forecastedvalues_ln=forecast(fitlnstock,h=3)
    > forecastedvalues_ln
    Point Forecast Lo 80 Hi 80 Lo 95 Hi 95
    2018 10.99061 10.93651 11.04472 10.90787 11.07336
    2019 10.94508 10.86857 11.02159 10.82807 11.06210
    2020 10.89955 10.80584 10.99326 10.75624 11.04286
    > plot(forecastedvalues_ln)
    > forecastedvaluesextracted=as.numeric(forecastedvalues_ln$mean)
    > finalforecastvalues=exp(forecastedvaluesextracted)
    > finalforecastvalues
    [1] 59314.82 56674.67 54152.04
    >
    > Box.test(fitlnstock$residuals,lag = 2,type = "Lyung-box")
    Error in match.arg(type) :
    'arg' should be one of “Box-Pierce”, “Ljung-Box”
    > Box.test(fitlnstock$residuals,lag = 2,type = "Lyung-box")
    Error in match.arg(type) :
    'arg' should be one of “Box-Pierce”, “Ljung-Box”
    > Box.test(fitlnstock$residuals,lag = 2,type = "Ljung-box")
    Error in match.arg(type) :
    'arg' should be one of “Box-Pierce”, “Ljung-Box”
    > Box.test(fitlnstock$residuals,lag = 2,type = "Ljung-Box")
    Box-Ljung test

    data: fitlnstock$residuals
    X-squared = 1.1, df = 2, p-value = 0.577

    > Box.test(fitlnstock$residuals,lag = 4,type = "Ljung-Box")

    Box-Ljung test
    data: fitlnstock$residuals
    X-squared = 2.0561, df = 4, p-value = 0.7254

    > Box.test(fitlnstock$residuals,lag = 6,type = "Ljung-Box")

    Box-Ljung test

    data: fitlnstock$residuals
    X-squared = 4.6182, df = 6, p-value = 0.5936

    >

    You might also like