Skip to main content

Create Data Model


The easiest way to learn how to write BlockML configurations is by looking at the Ecommerce example in the demo/p1 project.


You need some basic SQL skills or outside help for this step.

Click the repoName-branchName selector and select username-master.

This will bring you into edit mode on the BlockML page. This version control section may be helpful to you.

Think about which tables in your database have relationships with each other.

Create a .view file for each of these tables and describe the fields that might be of interest to analyze.

Next, create a .model file and specify which keys can be used to link your views.

To do this, hover the mouse pointer over the project name in the file tree on the left, click the ellipsis, click Create file.


You can have any folder structure in Mprove, but the filenames must be unique. We recommend using a short prefix in file names and using snake_case.

After clicking on Save, the BlockML is validated. If you make a mistake somewhere while editing BlockML, you will see it. If an error is made in the View, then you will also see errors in all related models, visualizations and dashboards.


First, troubleshoot the errors in the .view files - because they are causing errors in other related objects.

Once you have created .model file and your BlockML is valid, you can explore your model.