Displaying the general look of the time series (GAS)
tsdisplay(gas)
## AUTOPLOT
On the autoplot, it is noticable that there is significant rise in amplitude of time series data over time. Therefore it can be concluded that this series is multiplicative type
autoplot(gas)
The seasonal plot shows two meaningful factors.
There happens to exist a seasonality factor that is becoming more and more significant over years. Namely, each year gas production in Australia increases in the summer but the later the bigger is the difference between summer and winter
The amount of production despite month trends and outliers is generaly increasing over time. Therefore, we can observe a general increasing trend.
ggseasonplot(gas, year.labels = TRUE, year.labels.left = TRUE) + ylab("$") + ggtitle("Seasonal plot: australian monthly gas production")
ggsubseriesplot(gas) + ylab("$") + ggtitle("Seasonal plot: australian monthly gas production")
gglagplot(gas)
ggAcf(gas)
Decomposition of the time series show each factor of anaylyzing data.
The assumption of the general increasing trend is confirmed.
There is significant seasonal factor
decomposition <- decompose(gas,type="multiplicative")
autoplot(decomposition)
Function “which.max(gas)” shows the outlier which happens to be the 475th observation which is: November 1995 - 66600
which.max(gas)
## [1] 475