In Holistics, exploring data from a dataset is the starting point toward building a dashboard. Using datasets prepared by your analysts, you can slice and dice your data with a drag-and-drop interface, as well as combining different data models to answer complex questions.
How does data exploration work?
On the surface level, you interact with your data using a familiar drag-and-drop interface. Behind the scene, Holistics uses your interactions and the metadata in the modeling layer to generate queries, and then runs it against your database. When the result set is returned from the database, Holistics combines it with the visualization settings to render the result to your browser.
How to start exploring
Data exploration is one of our core focuses in Holistics, so we designed it to seamlessly blend in your analysis workflow.
Explore from a dataset
If you want to explore your data from scratch, navigate to the Datasets section under Reporting page, and click on the dataset you need. The exploration UI will appear with no field or conditions pre-selected:
Explore a Dashboard Widget
When you look at a dashboard and found a piece of data worth investigating, you can also explore data directly from a widget.
Simply right-click then click Explore, or click on More (...) -> Explore., and the exploration UI will appear with the fields, conditions and visualization settings pre-applied:
In general, the exploration UI is great for the less technical-inclined members. If you are well-versed in SQL, probably you would love to write your own queries in our SQL Editor.
In this section, we will introduce you to the basic concepts and functions of the Dataset Exploration view.
Dimensions, Measures, and Business Calculations
In general, you will be interacting with two types of field:
- Dimensions are non-aggregated data fields displayed in black. Dimensions can be of any data type.
- Measures are displayed in blue with a sigma icon, and they are used to aggregate your data (for example, counting, summing, averaging...). Measures always return a numerical result.
When opening a dataset, a list of dimensions and measures from its component data models will be displayed on the left.
On top of that, you can also use Business Calculation to derive new dimensions and measures from the available ones. To learn more about this feature, please refer to the documentation page.
To query a field, simply drag it into the Visualization Settings panel, and click Get Result:
Behind the scene, Holistics will use the metadata from the models and fields to construct a query to be run against your database.
You can also add a new field by clicking on the + Add field area and select the field from the dropdown list. From here you can also add new Business Calculations.
When exploring a dataset, there are three ways to aggregate your data:
- Using an aggregate function in the UI: Simply click on a dimension you already dragged in, and select a function from the list. This is mostly used at the initial exploration stage, when you want to understand the shape of your data.
- Using a Measure from the model: This is used when your aggregation logic is more complicated than simple counts or sums. The measure's formula is stored at Data Model layer, so it can be shared across multiple reports. For more details, please refer to the Custom Fields documentation page.
- Write a Business Calculations: With this, you can write custom aggregation logic on top of the current exploration result using Holistics Expression. However, the logic written here will not be available to other reports.
Visualizing data & Styling
Sometimes the numbers do not say it all, but a picture may worth a thousand words. Holistics provides you some of the commonly used chart types to visualize your data.
In each chart type, there will be further options to customize how you want to present your data. To learn more about charts in Holistics, please refer to the Visualizations documentation page.
To filter your data while exploring, drag a field into the Conditions area, select the comparison operator and specify the value to be applied.
Different field types will have different sets of operators, for example, conditions using a Date field will have operators that are specific for date comparison:
To learn more about basic filtering operations in Holistics, please visit the Filters documentation page.
For more advanced features, you can also check out:
The Format tab in the Visualization panel is where you further customize how you want to display data. For example, you can choose to display decimal numbers with one or two decimal points, or even display a date field as a string.
For more details, please refer to the Formatting Data page.
Sometimes, to speed up the process of exploring data, you only want to see a subset of your complete results from your query. You can do this in Holistics by using our Row Limit feature:
Without a set row limit, Holistics supports up to 1,000,000 rows for Data Exploration / Report Creation and 100,000 rows for Data Modeling Preview. When you set a row limit, Holistics will only display up to the number of rows you have set, and let you know about your incomplete results.
- The Row Limit is directly applied on your query.
- Our sort order is important in these situations: Holistics will first applies the limit, and then applies the sort.
Saving exploration as a report
When you are done with exploration, you can save the result as a report on a dashboard. Simply click on Save on the top right corner, enter the report title and search for a destination dashboard.
Note that any filtering condition you were applying while saving the report will be retained as a widget filter:
For more details on how to report your data, please refer to the Creating Dashboards documentation page.
Sharing your exploration
While exploring a dataset, you may want to share interesting tidbits that you found with others in your organization. There are two ways to do so:
Sharing your exploration state
For every change you make when exploring a dataset, Holistics will generate a unique URL, and you can give the link to your colleagues to continue with your exploration, provided that they have permission to the underlying datasets and models:
Sharing a static snapshot
On the other hand, if you only want to share the result of your exploration, you can export a snapshot of your current visualization and the underlying data:
For example, exploring both chart and data to an Excel file: