As you begin working with the prediction wizard, it's important to bear in mind the differences between a wizard template and a predictive model. A wizard template is used to generate a predictive model. A predictive model is a closed black box algorithm - once created it cannot be edited or changed. A wizard template can be changed at any time to rebuild an existing predictive model or to build a new one. When you save a wizard template and predictive model, they are both saved with the same name.
A simple analogy might be the distinction between a cake recipe and a baked cake. The wizard template is like a cake recipe which can be changed and improved at any time. The predictive model is more like an actual cake which cannot be remade after you remove it from the oven. If the cake is not tasty, you need to fix your recipe and bake a new one.
A wizard template is used to record the various selections you make in the wizard dialogs.
- A wizard template can be saved and then reloaded at a later time in order to preserve the selections that you make in the wizard dialogs.
- You can create many wizard templates, each with a unique name. You can also save a single wizard template multiple times with the same name, thus creating multiple versions of the same template. You can then choose which version will be the active one.
- A wizard template can be used with any report that is based on the same data model.
- When saving wizard templates, you can apply private or public access according to the standard BI Office security paradigm.
A predictive model can be used to predict new X values (hierarchy or measure) that are calculated by the R Engine.
- A predictive model cannot be edited. So if you want to tweak a predictive model, you need to actually create a new predictive model with your changes.
- You can create as many predictive models as you like.
- Each predictive model has a unique name. You cannot create multiple versions with the same name.
- A predictive model can be used with any report that is based on the same data model. It can also be used with a report based on a different data model, provided the model contains the EXACT same element names (cap sensitive) on which the predictive model is based (dimension, hierarchies, measures, etc.).
- When saving predictive models, you can apply private or public access according to the standard BI Office security paradigm.