4D Analytics

Creating a Model

Each model has two points associated with it. The first is designated as the ‘Dependent variable’ and the second is the ‘Independent variable’. To create a model, we first select the dependent variable in the 4D Analytics data tree.

Example:

Select the dependent point.

On the Model Details pane, click the ‘Add’ button the new model appears in the grid.

At this point, we drag and drop the independent point and drop on the model details.

The independent variable is entered into the details.

The following details can be edited by double clicking on the appropriate cell and entering appropriate values.

  • Name
  • Regression Type
  • Resolution
  • Start Date
  • End Date

It is recommended that the resolution is set to Day, when there is high resolution data for the selected points.

Including State

A third variable may be used as a state category. This is entirely optional. It allows the model to calculate results that fall into each defined category (e.g., 0 or 1, to indicate Store Open and Store Closed respectively). The state category is selectable from the graph (see later).

The state values are contained in a third point which has values covering the same period as the other points used in the model. It also must have data at the same resolution.

The State point is added in a similar manner to the Independent variable i.e. by dragging from the Tree and dropping into the State Point column.

Hovering over the name of a point in this column with the mouse will reveal its database ident in a tooltip.

Once a state Point has been assigned, it can easily be removed by right-clicking on the point name in the State Point column. The row will be selected and a popup menu will appear:

Select “Remove State Point” and the state point will be excluded from the model.