Creating a Databricks SQL Dashboard
1. Creating a Databricks SQL Dashboard
In this scenario, I am a data analyst at Wikipedia. Right now, you see I am in the Data Explorer and am looking at what tables I have at my disposal. I can see that I have data for different user click behaviors. One of our groups is curious about which pages our users are clicking on, and would like me to put together a dashboard for them to interact with and dive deeper into the data. So, to start things off, I will navigate over to the SQL Editor view. In this view, I am still able to see the schema, tables, and columns I can access, but I can also start to write different queries. In this first query, I want to get all of my data and use visuals to better understand user behavior. I created two visuals for this first query, one that shows the top pages, and another that shows a histogram of number of clicks on each page. Now let us create a new visual. I want to create a pie chart, called Referrer Distribution, so we can see the distribution of click count for each referring page as a percentage of the total. In the x-axis I’ll put the referrer column, and in the y-axis Ill select the click_count field. I can configure the visual to look how I would like, and can save the visual. My next query will be more specific, as my end users often will want to look at a particular topic. For this query, I will take a look at a different table and can provide a parameter in the filtering criteria. In this case, I create a text parameter called “Page Keyword”, which will allow my query to find all similar text from the dataset. The corresponding visual will update its value based on the parameter you input. Finally, we can put all of these visuals together into one dashboard. I will go to the Dashboard section and click on Wikipedia Dashboard. When I click the Add dropdown, I can start to populate my dashboard with the visuals I created as part of my previous queries. I can also add in any text boxes that I want, such as a title box for my dashboard. I’ll create a title that says “Wikipedia Clickstream Dashboard”, and will add in my histogram chart and the counter visuals from this video. As we expect, I can have each visual side by side, and my parameter has made it all the way to the dashboard, so my end user can update what topic they would like to analyze. Since we added a new pie visual, I can edit my dashboard, select the correct visual from the query, and add that to my dashboard. This was a quick example of how to use queries, visuals, and dashboards to perform your analyses.2. Let's practice!
Create Your Free Account
or
By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.