Use the canadian_gas data from the fpp3 package in R and perform the following tasks. Include explanations, wherever needed.

1. Subset the above data and keep observations from March 1990 to February 2005. Describe the data set.

2. Plot the data using autoplot(), gg_subseries(), and gg_season(). Describe the presence (or absence) of seasonality in the data.

3. Perform an STL decomposition of the data. Plot the seasonal component from STL decomposition using gg_season(). Does the seasonal shape change over time? Explain.

I keep getting errors. Please explain the steps taken to set it up, thanks.