Data Preparation¶
The Data preparation pill provides a visual way to query a metric view, intended for business users with little knowledge of the view schema or VQL. It operates on a what you see is what you get principle.
Data preparation screen to build queries by modifying real-time data¶
Unlike other methods, this approach begins by retrieving data from the view. Instead of loading the entire data set, a subset is obtained that serves as a representative sample of the view.
Note
Unlike standard views, metric views aggregate data dynamically and cannot be sampled. Therefore, to use data preparation, live data has to be retrieved directly from the view.
If data preparation is configured to use a sample of the dataset, an exception must be enabled for metric views in the data preparation settings so they can be queried live.
To build the query, you have access to several actions that can be applied to modify the data. Each action processes the live data in real time, allowing you to immediately visualize its effect and assess whether it meets your objective. Additional actions can then be applied iteratively until the desired result is achieved.
Once all actions have been applied and the final query has been defined, it is ready to be saved or executed.
For comprehensive instructions on how to load data, apply filters, and use transformations, please refer to the data preparation documentation page for views.
Next, we will outline the specifics on how to start a data preparation session for metric views.
Loading Data¶
Each data preparation session begins in the Columns section. The metric view fields are organized into Dimensions and Metrics.
To define the fields that will be included in the query, drag the desired fields from either panel to the Output Column panel. If you drag a dimension, all of its attributes will be added to your query.
Because metric views are specifically designed to aggregate data, a query must always include at least one dimension in the output columns. It is valid to execute a query using only dimension fields but it is not allowed to execute a query with only metric fields.
Once you have selected the required fields, click Next to continue. The system will redirect you to the Definition section, where a preliminary preview of the query results is shown based on the selected output columns.
Take into account that the Columns section is always available and you can go back there whenever you want. To discard your data preparation session and start from the beginning, click the Start over button.
Wizard to select the fields that participate in the query¶
Visualizing the Result¶
When you reach the final stage of the query definition, you can easily explore your data visually. In the Results section, a Visualize button is available in the toolbar.
Clicking this button automatically triggers the Denodo Chatbot Assistant. The assistant analyzes the structure of the query and dynamically generates an appropriate chart that best represents the dataset.
Visualize button available in the results section¶
Denodo Chatbot Assistant displaying the generated chart¶
Note
The Visualize button is only available if the Chatbot has been enabled. To learn how to activate and configure this feature, please refer to the Denodo Assistant Chatbot Configuration documentation page.