2. Getting the picture - graphing the raw data.
Graphing Raw data and Discussing Features
The first thing to do is produce a graph of the raw data.
This gives a good idea of the 'story' of the data
What can we see?
Trend - long term movement of the data
Seasonality - repeated and regular cycles
Longer Cycles - repeated patterns in the data that occur over longer periods
Variation :
random changes - 'noise'
one off events (outliers or spikes)
A line graph is used to graph time series. Correct labelling of axis and time intervals is important.
How do we describe features of a time series graph?
When describing features of time series graphs specific values must be referred to.
Multiplicative & Additive Models