What is ACF in ARIMA in R

Recipe Objective

What is ACF in ARIMA?

Time series is a statistical technique that deals with time series data or trend analysis. Time series data means the data is collected over a period of time/ intervals. Time series data helps us with making forecasting based on the previously collected data. ARIMA is a time series forecasting method where AR stands for Autoregressive and MA stands for Moving Average. The AR model is based on the assumption that the data collected is stationary and univariate, and the forecast corresponds to a linear combination of previous data of the variables. In order to calculate the auto-correlation between an observation of a time series data, ACF (Autocorrelation function is used). It calculates the lag values for AR and MA in ARIMA modelling. This recipe demonstrates an example of ACF in ARIMA.

Learn About the Application of ARCH and GARCH models in Real-World

Step 1 - Install required package

install.packages('forecast')
library(forecast)

Step 2 - Generate random time series data

# Get the data points in form of a R vector.
rain <- c(987,1025,978,774,1563,569,1456,789,1479,566,1563,1698)
# Convert it to a time series object.
rain_ts <- ts(rain,start = c(2020,1),frequency = 12)
# Print the timeseries data.
print(rain_ts)
plot(rain_ts, main = "Time series data")
summary(rain_ts)

Step 3 - Plot a trend line

plot(rain_ts)
abline(reg=lm(rain_ts~time(rain_ts)))

Step 4 - Build a model using arima()

model <- arima(rain_ts, order = c(1,0,0)) model

Step 5 - Plot a acf()

acf(ts(diff(log10(rain_ts))),main='ACF for rain_ts')