In order to make a chart, often times you will first need to summarize your data by rolling up the rows based on a common value. For example, I have a dataset that lists the funds for unclaimed bank accounts, and there are many accounts that belong to the same bank. I want to find out the total unclaimed funds at each bank and make a pie chart to visualize my results.

Here’s how I do that:

1. When you have your dataset open, click on the blue “Filter” icon on the right.
2. In the Filter right side bar, click on “Sort & Roll-up” and then check the box for “Roll-Ups and Drill Downs” to get the menu, seen in the screenshot below:
3. Under “Group By” select the column name that has the values you want to summarize. In my case, I want to group by “bank_name.”
4. Under “Roll-Up” Select the values you want to roll-up, and “Function” of how you want to roll them up. In may case I want to roll-up “Balance” and I want the function to be sum, because I want to know the total balance for each bank. If I wanted to count the number of accounts per bank, I would want to select the function “count.”
5. Click “Apply” and you’ve successfully created your first roll-up! Below is what my new view of the dataset looks like:
6. Now, to make a pie chart of my filtered data, I follow the steps to make a pie chart, and get the chart below. Looks like the Canadian Imperial Bank of Commerce has a lot of unclaimed money!