A Data Model is a non-physical but logical element, which allows to define a context of analysis and is composed of one or several Datasets and their respective Relationships.
Once a data model has been defined, it is available to carry out the data analysis it allows to the whole Organization, following the concept of "define once and use many".
The option "Data Models" is available on the left side menu of the Data Center, which gives access to the list of data models that your organization has defined in Biuwer.
From this list you can see the different Data Models that your Organization can use in any data card to make data queries and perform the following operations:
Filter data models according to their name and data connection
Directly access the model detail by clicking on its name.
From the context menu of each model, see the model detail, edit it or delete it.
Add a new model.
As in the case of Datasets, we also distinguish between managed and not managed Models. By clicking on the "Add" button in the list of models, a data model creation dialog appears. By default the option "Not Managed" is selected in which it is necessary to select a data connection. A not managed model will allow to include in it any dataset associated with a specified connection, because the query will be executed in the database engine associated with the connection.
If, on the other hand, you want to create a "Managed" model, it is not necessary to include a connection when you select this option, because Biuwer manages it automatically for your organization as a CDW (Cloud Data Warehouse).
Managed models can only include managed datasets. This is interesting because relationships can be modeled between data of different types uploaded to Biuwer, obtained through the CSV or Excel file importer, connections to external APIs or through the Biuwer data preparation module.
We have access to the list of Datasets that are part of the model from its detail page.
Using the "Add" button we can include new datasets. From the list itself, by clicking on the red cross you can remove a dataset from the model.
Clicking on the "Relations" tab shows the list of relations included in the model.
A relationship links two datasets using one or more fields, using an operator (equality, inequality, greater or lesser) and a relationship type:
Inner: includes all records when there is at least one match in both datasets.
Left Outer: includes all records of the left dataset, and the matching records of the right dataset.
Right Outer: includes all records of the right dataset, and the matching records of the left dataset.
Full Outer: includes all records of the left and right datasets.