Tableau charts Archives - Mindaugas Jasas

Tableau Heat Map Tutorial: 3 examples

October 20, 2019 MindaugasJ Leave a comment

What is a Tableau Heat Map?

Heat map in Tableau is a data visualization type for which you need to have one or more dimensions and 1 or 2 measures. In these heat maps, Tableau is displaying a table consisting of many squares. Tableau can represent these squares in different sizes and colors. You can choose measures that will reveal data by size, color, or both.

How do you create a Heat map?

To create a heat map, you need to have a data set with at least in dimension and one measure. Ideally, you must have two dimensions, of which one should be a date and one or two measures. When you select your data source, follow these steps:

Select your dimensions and measures when holding a CRTL button.
Press Show Me button in the top-right corner and select Heat Maps to create our Tableau heat map.

What are the best practices to use heat maps?

To use a heat map in Tableau is a go-to practice when you have some measure that you can analyze by timeline or sub-categorical dimensions that divides the sum of that measure. If you have too many rows and columns in your heat map, then you can add filters to make finding insights easier.

Example 1. Determining which region has the most sales

We will use Superstore Sales data for this first example. Firstly, connect to our data by selecting the Orders sheet as our data source. Then we immediately go to our Tableau Sheet 1 and renaming it to what we want to find in it. In this instance, change Sheet 1 to Product Subcategory Sales by Region. As a result, we now see the modified sheet header and can go to the next steps that will create a heat map.

As you can see in the following video, we are selecting Product Subcategory, Region, and Sales. Then, we choose Heat Maps from Show Me list. As a result, we have our first heat map that can be formatted to look better. We switch column with header because instantly, Tableau put it in alphabetical order. Then, change Region header alignment to vertical.

Tweaking the heat map

Even though a Tableau heat map is intact, we still not have the best version of the chart in front of us. Indeed, there are a few things that we can improve here. First, increase the size of a heat map so it would look more appealing. Second, switch SUM(Sales) pill in the marks field from Size to Color and choose whatever color palette you like.

Now we can see that most sales were in Ontario, Prairie, and West regions. Also, it’s clear there were no Office Machines sales in the Nunavut region as it’s rectangle is white.

Moreover, we can use the second measure in the heat map to display both different sizes and colors for rectangles. Let’s change SUM(Sales) pill to size and drag Profit measure to colors field. As a result, we can understand which subcategories by region have the most sales (width of rectangles) and profit (color of boxes). With this addition, we can recognize negative profit cases. For example, we can see that Bookcases our store sold with negative profit in Yukon, Northwest Territories, Ontario, and Quebec.

Example 2. Best App Store games 2008-2019

Let’s open the 17k app store games database from Kaggle. Steps that I use to create a Tableau heat map are in this video:

Using various filters, you can determine multiple app store games that have the biggest User Rating Count. As we know that the best games get rated more, the big User Rating count can indicate the best seller games in the Apple app store. If you want to know the best Role-Playing games, filter this genre and set the minimum User Rating count such that optimal number games could appear in a worksheet. For example, after placing the minimum User Rating to 20.000, you can see a Tableau heat map consisting of probably the best Role Playing Games in App Store history. Tableau Public worksheet is shared below.

Example 3. CO2 Emissions By country in the Tableau Heat Map

This example is meant to show countries with the most significant CO2 emission and also changes in the emission numbers in the timeline. You can download the CO2 Emission data here. I am not going into further aspects of what steps I am taking to build a heat map in this case. Just watch the video below and digest it.

Conclusion

To summarize, a Tableau heat map is an excellent alternative over traditional excel tables with numbers. In numerous circumstances, it is easier to find patterns using the size and color of cells over numbers.

Tableau Tableau charts, Tableau examples, Tableau heat map, Tableau tutorial

Tableau pie chart: full tutorial

October 16, 2019 MindaugasJ Leave a comment

The pie chart concept comes from old times. The pie distribution problem: how to divide a pie into X equal parts for X people. Nowadays people use pie charts to asses what components have the biggest share in some totality. In Tableau visualization, a pie chart is one of the most popular charts. For business applications, this chart is most appropriate to demonstrate profit by country, sales by client or expenditures by a company branch.

For this tutorial I am going to use the same Superstore Sample data set that I’ve used in Tableau bar chart tutorial.

Creating a pie chart

To create a pie chart in Tableau, we need to have 1 and more dimensions and 1 or 2 measures. In this tutorial, I drag Sales to rows and Product Category to columns. However, a bar chart was created automatically at first. After that, we press Show Me and select pie charts. By doing so, we create a pie chart.

Other creation method

If we want to avoid creating a bar chart, we need to follow this instruction:

Select Pie in the Marks field;
Drag Sales (a measure that you want to asses) to Size and Angle buttons;
Drag Product Category (a dimension by which you want to divide your measure) to Color button.

Formatting a Tableau pie chart

Changing size

Changing a size for a Tableau pie chart is one of the most annoying tasks. We have to move our pointer to the right next to a pie chart. As our pointer changes it’s appearance to a two way arrow, we can press, hold and drag it to the right to widen chart size limiter. Also do the same with the bottom limiter to attain a bigger chart.

Display labels

To display labels, drag Product Category dimension to a Label button. If you want your labels to be inside of a chart, you can click on those labels and drag to a center of an angle or wherever you want. To see what percentage of a whole share each category has, drag Sales to a label button. After that, select the SUM(Sales) label mark and add Quick table calculation – percent of total.

Tableau Donut chart

Donut chart is a variation of a Tableau pie chart. To create it, we need to:

Drag Number of Record measure to rows column twice;
Set measure to a minimum for both of those Number of Records marks that are in rows field;
Then we select Dual-axis in the second Number of Records mark;
In the marks field of MIN(Number of Records) 2, remove Product Category from colors;
Then, press the labels button and mark out “Show mark labels” field;
Select white color or whatever color that suits your Tableau worksheet;
Press on Size button and shrink your additional pie chart that will serve as a donut hole;
Select one of the axes and mark out a header.

Tableau Advanced: Gauge KPI chart

Business people love seeing Key Performance Indexes (KPI) expressed in Donut or Gauge charts. There are no pre-built Gauge chart in Tableau, so users are using their creativity to build it themselves. A Gauge chart in Tableau is usually just a half of the Donut Pie chart.

You can find a lot of tutorials over the internet of how to create Tableau KPI Gauge charts of different style:

The sad part of creating a gauge chart using those tutorials is a lack of clarity. Great attention to detail is required to understand the concepts used in those methods. Additionally, you have to ascertain how to implement those methods with your data. Luckily, if you’ll take your time to comprehend Gauge creating concepts, completing your analytic work tasks will be easy.

Creation steps for Gauge chart

I am taking Gauge creation concept from Vizartpandey.com blogpost about creating semi donut chart, because it is as simple as possible.

First, we have to understand that default pie chart angle order can’t be used to represent a gauge chart. You can see why in this picture:

1. Creating supporting data and calculated fields

To change pie chart angle order so it would suit gauge chart, we need to create supporting data that will make new angles.

Sections

Zero_To_Value

Value_To_50

After_50

Remainder

Bottom Half

We add column ‘Number of Records’ and enter ‘1’ for every section row. Also, we do the same with whatever data for which we want to create gauge charts. You can do this addition in Excel easily. This will allow those rows cross-join, when doing FULL-OUTER join on ‘Number of Records’=’1’ in Tableau Data Source window. As a result, we will have 5 different sections for each row of our data-set.

We are going to create a gauge chart that will represent Free Throw percentage of NBA players. Lets use 2018-2019 season stats from https://www.basketball-reference.com.

Now we create calculated fields.

[Max %]– Free Throw Percentage Value based on each player

MAX({FIXED [Player]:MAX([FT%])})

[Arc Angle] – Logical calculations needed to transform our percentage value so that it will be represented in upper half of a circle.

CASE ATTR([Sections])
WHEN “Zero_To_Value” THEN IF ([Max %])<= 0.5 THEN ([Max %])/2 ELSE 0.25 END
WHEN “Value_To_50” THEN IF ([Max %])<= 0.5 THEN (0.5-[Max %])/2 ELSE 0 END
WHEN “After_50” THEN IF ([Max %])> 0.5 THEN (([Max %])-0.5)/2 ELSE 0 END
WHEN “Remainder” THEN IF ([Max %])<= 0.5 THEN 0.25 ELSE (1-([Max %]))/2 END
WHEN “Bottom Half” THEN 0.5
END

2. Build a Doughnut

To build a Doughnut chart that will be used to create a Gauge chart, we:

Drag ‘Number of Records‘ to Rows field twice;
Set the aggregation of it to minimum;
Right-click on second pill of ‘MIN([Number of Records])’ and select Dual Axis.
In the marks section, select first ‘MIN([Number of Records])’ mark and set the chart type to Pie. For the same mark, drag Sections to Color and Arc Angle to Angle.
Then select second ‘MIN([Number of Records])’ and set the chart type to Circle. Add Max % to label of this circle and drag this label higher so it will be showing in the top part. Change circle color to white and reduce size of the circle. Now we see a doughnut.

3. Finish the Gauge chart

To finish Gauge chart we finish these steps:

Sort Sections manually in this manner (Also assign following colors):

Sections

After_50 –Yellow

Remainder –Grey

Bottom Half –Any Color (Bottom will be hidden after we’ll fix the axis)

Zero_To_Value –Yellow

Value_To_50 –Yellow
Drag Players to Filters card. Press Analysis –> Filters and add Players filter to the right panel. Change Players filter so it will be single value drop-down;
Edit axis so it’s range will be fixed from 1 to 2;
Remove axis headers so we will be left with a Gauge chart.

You can copy numerous of these Gauge charts to the Tableau dashboard. In this case I am duplicating my sheet and moving both Gauges to the Dashboard to compare FT% of two players. Check my Tableau Public dashboard below:

A lovely chart indeed

The pie chart is commonly used in Tableau and for a reason. It’s one of the best Tableau charts you can use. It presents data accurately, shows both proportions and values, and is very easy to interpret. Notably, the classical pie chart is easy to make. Additionally, there are more complicated variations of the Tableau pie chart like Donut chart, Gauge chart and Pie chart on Map. These charts can be used to represent data in BI visualization reports more clearly.

Tableau Tableau charts, Tableau pie chart

Tableau bar chart tutorial

October 13, 2019 MindaugasJ One comment

The fastest way to compare your data is creating Tableau bar chart. In this tutorial, I choose to open Superstore Sales file that I’ve mentioned in my datasets for analysis page. After opening excel file, I choose Orders sheet and drag to data source field.

Selecting measures and dimensions

After connecting with the Super Store Sales file, we see these measures and dimensions: –>

Let’s think what we want to see in our bar chart.

Profit is one of the main KPI’s in business, we are choosing that measure and dragging to rows field. Furthermore, to see what product categories and sub-categories are generating most profit, we need to drag those dimensions to the columns field.

Polishing the Tableau Bar Chart

The bar chart is made now, but we have to make it more beautiful and easier to read. Here are the steps that we follow to upgrade our chart:

Move Office supplies category to the left, because it has most sub-categories. Also move technology category in front of the office supplies, because technology bars are much higher. This way it will look better.
Add profit numbers to bars by dragging Profit measure to label field
Now we see that Telephones and Communication category label is not there. Number that would be used for a label is too big, so we decrease text size from 9 to 8.
Sort sub-categories by profit (descending).
Add colors to Tableau bar chart to make things further clear. Drag Profit measure to color field. I chose to use red-green diverging for this example. Also, I have marked Use Full Color Range box so that the worst sub-category will be visualized with the richest red in the range.

Final chart view

After finishing our bar chart we can see which categories and sub-categories makes good profit numbers. We see that “the Store” loses money by selling Bookcases and Tables. In a real-world situation, middle managers would be required to report what were the reasons for that. Additionally, as analysts, we can drill-down into subcategories and try to find those reasons in our data.

Tableau Tableau charts, Tableau tutorial