Leveraging Advanced Analytics

This chapter focuses on advanced self-service analytics. Self-service analytics can be seen as a form of business intelligence, where people in a business are encouraged to execute queries on datasets themselves, instead of placing requests for queries in a backlog with their IT team. Then, query analysis can be done, which should lead to more insights and data-driven decision-making. But how do you start creating useful self-service dashboards if it’s your first time doing so? How do you go from a dataset to a product? Have you ever asked yourself how other people start working on a dashboard, how they clean data, and how they come up with a dashboard design? If so, this is the right chapter for you! I want to share three use cases with you, written as a train of thought in order to give you an idea about how I work. Please note that this is just my personal experience; there are many different ways to reach your goal.

We will cover the following...

Visualizing world indices correlations

Imagine you are working on the world indices dataset and your line manager gives you the following task:

Create a dashboard for me in which I can easily spot all correlated world indices and their distribution. I need it by tomorrow morning.

Now, take a few minutes before you continue reading and think about how you would tackle this task. The dataset contains 67 columns with various indices, like birth registrations or emission values, exports and imports, and forest areas, divided into 188 rows, where each row represents one country.

Write down your planned steps, open the workbook related to this chapter from https://public.tableau.com/profile/marleen.meier, and follow your steps; time it in order to get a better feel for time estimates when working with Tableau. This way, you can make sure that you can deliver on time and manage expectations if you are ever asked how long it will take to build a certain dashboard.

Plotting...

Geo-spatial analytics with Chicago traffic violations

It’s Wednesday morning; your manager comes into your office wanting to check the red-light violations in the last year in Chicago. They ask if you can build a dashboard for that purpose. In particular, you’re asked to highlight where the most violations happen and whether there is an overall trend in Chicago traffic light violations over the last few years. You are given two datasets, one with the camera locations and one with the violations, and are told that the dashboard is needed within the next hour. What do you do?

Before you continue reading, think about how you would approach this problem. Take five minutes, think about the steps you would take, and sketch a dashboard design.

The following is an overview of how I would do it:

1. Open the datasets in Tableau Prep Builder.
2. Join the two datasets.
3. Clean the data if needed.
4. Open the output in Tableau.
5. Use a map to visualize...

Building a map of intersections

I continue by opening the violations dataset in Tableau:

1. Longitude and latitude values are not automatically recognized, so I have to change both to Number (decimal) by clicking on the data type icon:

Figure 12.41: Change data type

2. Then, I change the Longitude and Latitude fields to Measure by clicking on the drop-down arrow on the field as shown in the preceding figure and selecting Convert to Measure.
3. Now I can click on the data type icon again and change the two fields to Latitude and Longitude:

Figure 12.42: Changing the geographic role

4. By dragging Longitude to Columns, Latitude to Rows, and Intersection to the Text shelf, I visualize the red-light locations—at least the ones that have ever had a violation:

Figure 12.43: Intersections in Chicago

5. The name of the worksheet will be Intersection, and since I am looking at violations, I change the color...

Extending geo-spatial analytics with distance measures

Our last use case is also geo-spatial analysis on the same Chicago traffic dataset, but this time, we will add another component. We will look to rent a new place but with the requirement that there are no more than n intersections in a radius of x and Navy Pier should be at most y miles away. The variables n, x, and y should be interactive in order for us to make changes and have a very flexible dashboard experience. The questions to ask about this task are:

• How can we add any given location in Chicago to our dataset? It is currently only showing intersections and violations.
• How can we make the n, x, and y variables?
• How can we add a radius indicator to any given point on the map?
• How can we measure the distance between two variable points?

All those questions will be answered in the following steps:

1. Go back to the workbook related to this chapter.
2. Right-click on the worksheet...

Summary

In this chapter, we explored the fascinating world of advanced self-service analytics. We began by understanding the concept of self-service analytics as a form of business intelligence, empowering individuals to directly query datasets and unlock valuable insights. Throughout our journey, we uncovered three captivating use cases that showcased the power of self-service analytics. We unraveled the intricate correlations between world indices, gaining a deeper understanding of global market dynamics. We then delved into the realm of Chicago traffic violations, using geo-spatial analytics to uncover patterns and potential areas for improvement. Lastly, we extended our geo-spatial analysis, utilizing distance measures to determine the optimal radius for housing locations based on key variables.

Amidst these diverse scenarios, one key lesson resonated throughout: a well-structured approach is paramount when embarking on a self-service analytics project. By carefully planning...