Knowledge Base & Community Wiki
Time Series Decomposition – Tutorial I
What is Time Series Decomposition – Decomposition procedures are used in time series to describe the trend and seasonal factors in a time series. More extensive decomposition’s might also include long-run cycles, holiday effects, and day of week effects, so on. Time Series Decomposition in VisualizeIT breaks down the original Time Series into its trend and seasonal decomposition.
One of the main objectives for a decomposition is to estimate seasonal effects that can be used to create and present seasonally adjusted values. A seasonally adjusted value removes the seasonal effect from a value so that trends can be seen more clearly. For example, in many regions of the U.S. unemployment tends to decrease in the summer due to increased employment in agricultural areas. Thus a drop in the unemployment rate in June compared to May doesn’t necessarily indicate that there’s a trend toward lower unemployment in the country. To see whether there is a real trend, we should adjust for the fact that unemployment is always lower in June than in May.
The following two structures are considered for basic decomposition models:
- Additive: Xt = Trend + Seasonal + Residual
- Multiplicative: Xt = Trend * Seasonal * Residual
The “Residual” term is often called “Irregular” in decompositions.
What is the capability provided in VisualizeIT as part of Time Series Decomposition – The Time Series Decomposition capability within VisualizeIT allows you Decompose data for a given Application / Data Dimension into the following components –
The Time Series Decomposition capability within VisualizeIT should be used to understand the Trends and Seasonality within your data. We would encourage use of the Time Series Decomposition capability to decompose the data into Trends, Seasonality and Residue once you’ve have used Time Series EDA (Exploratory Data Analysis) to slice, dice your data and understand the data sets key characteristics.
A precursor to using the Time Series Decomposition capability within VisualizeIT is to understand the “Periodicity” within your data. The “Periodicity” we are referring to is the “Seasonal Periodicity” that might or might not be present in your series. Identifying “Seasonal Periodicity” or “Periodicity” is the process of identifying this repetitive pattern that exists in your data. Keep in mind that “Seasonal Periodicity” is different from “Trend” within the data. This can be explained as follows i.e. Data shows a long term growth “Trend” with monthly “Periodicity” showing rise / falls that occur every month and possibly are more pronounced on some months of the year as compared to others.
Not all data sets will clearly display “Periodicity” and this is perfectly understandable. We would highly recommend that you explore your data using the Time Series EDA capability within VisualizeIT. Slice your data, dice your data and explore your Time Series data from different angles for different time periods to understand what the “Periodicity” is in your data set is using the Time Series EDA capability within VisualizeIT.
A per-requisite for using the Time Series EDA functionality within VisualizeIT is that you have uploaded data for the given Application / Data Dimension using the Data Management functionality within VisualizeIT. With the data uploaded you are then able to use the Time Series Decomposition functionality to explore your data.
Let’s start by taking a look at the approach to using the Time Series Decomposition functionality within VisualizeIT. To get started you would need to log-in to VisualizeIT, select the Statistical Modelling link from the Left Hand Side menu and then select Time Series Decomposition. You are then presented with the Time Series Decomposition form. Step 1 of the Time Series Decomposition form (above) requires you to select an Application including an associated Data Dimension.
Step 2 of the Time Series Decomposition form (above) requires you to select the time period i.e. Start Time, End Time and the Roll-Up required for the given Application / Data Dimension.
Step 3 of the Time Series Decomposition form (above) requires you to select the “Periodicity” for the given data set. This is the final step in the input parameter selection process. You then hit the “Visualize Data” button which then sends these input parameters to the VisualizeIT modelling engine and provides a set of graphical results which includes graphs displaying the Trend, Seasonality and Residual data for the given Application / Data Dimension.
The resulting graphs displayed provide a view of the Trend, Seasonality including the Residual data for the given Application / Data Dimension for the given period. The selection of “Periodicity” for the selected time period will influence the detected Trend and Seasonality.
Conclusion – The Time Series Decomposition capability within VisualizeIT provides you the ability to decompose the data set into its Trend and Seasonality. Your choice of Periodicity will influence the nature of Trend and Seasonality detected within the given data set. We would recommend use of the Time Series EDA (Exploratory Data Analysis) capability to understand the data and its various characteristics before using Time Series Decomposition to decompose your data into its Trend, Seasonality and Residue.
Modelling Solution: VisualizeIT offers access to a bunch of Analytical Models, Statistical Models and Simulation Models for purposes of Visualization, Modelling & Forecasting. Access to all the Analytical (Mathematical) models is free. We recommend you try out the Analytical models at VisualizeIT which are free to use and drop us a note with your suggestions, input and comments. You can access the VisualizeIT website here and the VisualizeIT modelling solution here –VisualizeIT.