Exploring data with Metabase's data browser

In this article, we’ll show how you can use Metabase’s Data Browser and other reference tools to learn about your datasets, as well as how administrators can curate information in Metabase about their data.

All the examples in this article use the Sample Dataset that’s automatically installed with your Metabase instance. If you want to follow along with your own dataset, first connect your database to your Metabase instance by following the instructions in Managing Databases.

Users can browse data in these two areas of the application:

Browse Data makes it easy to navigate by database or table name.
The Data Reference area provides detailed information about every table.

Admins additionally have access to these areas within the Metabase Admin section:

The Databases page gives a high level view of your databases.
The Data Model tab allows admins to edit table and column metadata.

Browse Data

Click on the Browse Data section at the top right of the navigation bar to display all databases currently connected to your Metabase instance. By clicking on a particular database, you can quickly navigate to different tables and view the information stored in each.

As depicted in Figure 1, when you hover over tables, three icons appear:

The yellow lightning bolt will create an X-ray, as discussed in the next section.
The Gray book will take you to the you to the Table’s Data Reference page.
The Information symbol will show a tooltip with the table’s description.

To view the rows in the table, click on the table’s name.

Figure 1. Navigating through the browse data screen to easily access information.

Lightning bolts create X-rays

The Browse Data section allows you to navigate databases and tables, view metadata about those tables, and create X-rays. X-rays are a way to autogenerate questions and explorations of your data that you can customize to suit your needs. At Metabase, we’ve seen quite a few dashboards, so we have a good idea of the kinds of things that folks often want to see, and we’ve designed X-rays to give people quick insights into their tables. Figure 2 is an X-ray that was automatically created for the People table.

Figure 2. Example of an X-ray based on the people table.

To create an X-ray in Browse Data, hover over a table and click on the yellow lightning bolt (Figure 3).

Figure 3. Hovering over a table in the <strong>Browse Data</strong> section, revealing a lightning icon and a book icon with the text 'X-ray this table'. — Figure 3. Hovering over a table in the **Browse Data** section, revealing a lightning icon and a book icon with the text 'X-ray this table'.

Anywhere these lightning bolts appear is a place from which you can create an X-ray.

Data Reference pages

The Data Reference section is a great place to look if you’re not sure whether a database has information that’s relevant to you. As an admin, you can use the space to clarify which users should be paying attention to the database, and which would find it unhelpful.

From any of the pages in the Browse Data section, select the book button with Learn about our data in the upper right corner. You are now in the Data Reference area; it’s essentially a data dictionary.

As you can see in Figure 4, this page includes database names and descriptions, as well as information about segments and metrics. Segments are filters that you can easily reference in the query builder, and metrics are an easy way to refer to a computed number (for example, revenue).

Figure 4. Landing page for the data reference section. Three tabs on the left that say Segments, Metrics, and Our Data.

Let’s dive into a database. Once we click into our Sample Database, two tabs appear on the left side of the screen (Figure 5).

The Details tab contains metadata about this database. As you can see in Figure 5, the Details tab features three sections that you can use to provide information about this dataset to your users:

General description
Why this database is interesting
Things to be aware of about this database

Figure 5. Data reference screen for the sample dataset. Not only can you see the dataset's description, you can also see why the dataset might be useful or edit the page's information.

If no description exists for those areas, people will see semi-transparent text to indicate the field has not been filled out by an admin. Admins have the option to click on the Edit button in the upper right hand corner to update this information (which they should).

The Tables tab displays the table names and descriptions. You can click on a table to view a Details tab of the table, as well as view the Fields in this table. You can view a list of Questions about this table (provided you have permission to view those questions), and have the option to create an X-ray of the table.

The Details tab suggests useful questions to ask the table at the bottom of the page (Figure 6).

Figure 6. The details tab of the Orders table.

Admins can edit field names and types in the Fields in this table tab (Figure 7).

Figure 7. The fields tab of the Orders table from an admin's perspective (with the Edit button in the upper right.

Access Metabase Admin

Clicking on Admin from the gear icon in the top right of Metabase will bring you to the Admin panel (Figure 8).

Figure 8. The landing page for Metabase Admin. The blank circles next to the items in the list indicate that nothing has been set up yet.

From the Admin panel, you can add, update, and remove databases, as well as edit metadata about your data.

The Databases page

The Databases page in the Admin panel displays connection information about your databases:

The database type
How Metabase is connected to your Metabase instance
Sync settings

Figure 9. Navigating within Metabase Admin to the overview page for the Sample Dataset.

Metabase does a lightweight sync every hour to keep your in-app data current, but you can use this page to manually sync your database, or manage sync frequency.

Edit metadata in Data Model

By initially organizing your metadata, you can save your users a lot of time and frustration. Picking clear names and adding useful descriptions for context will help users improve their analysis. Metabase can automatically try to create human-readable names of your tables and columns for you, but if we miss the mark, you can always disable the Friendly Table and Field Names feature in the Admin panel.

To make changes to your metadata in Metabase, visit the Data Model section in the Admin Panel. The Data Model page displays options to edit metadata for the database, tables, and columns. For example, you can edit a column’s name, visibility, type, and description. You can also remap foreign keys to give human readable names to foreign key columns!

Here are some ways editing metadata can make life better for your users:

When column names are confusing, you can change their names or add a description.
If you have address columns, you can hide them from your users to protect privacy.
You can pick your preferred filter interface from three options (search box, list of values, or plain input box).

Perhaps the most important piece of metadata you can change is the data type (far right of the screen in Figure 10). As depicted in Figure 8 below, there is a long list of data types to choose from. Selecting the correct data type for a column can connect information across multiple tables, and give context to Metabase so it can choose visualizations appropriate for your data types. For example, once you’ve accurately identified latitude and longitude columns in your table, you will be able to use map visualizations.

Figure 10. The data model for the Orders table, featuring editable metadata fields.