Give a chestnut! Tableau Tips (39): Cleverly Making Classifications of Sankeydiagram

Before, we shared the practice of Sankitu:Give a chestnut! Tableau Tips (25): Learn to do beautiful Sankeydiagram.

We know that Sangkitu is a specific type of flow chart. The width of the extended branches in the figure corresponds to the size of the data flow. It is usually applied to the visual analysis of energy, material composition, finance, retail and other data.

But have you seen this kind of Sanki map below? How is it different from the above-mentioned Sankitu?

In fact, the brand's Viz refers to the “energy efficiency map of steam engines” by Mathhew Henry Phineas Riall Sankey in 1898, which is an extension of “Sangkitu”.

The two approaches are different.They are different because they are fundamentally different for the presentation of data analysis:The trunk part of the brand tree (the line of Sanjitu) does not reflect the value, but only plays the role of classification.

So, how is the classification of the brand tree similar to the Sanji map? 

In this issue of "Give a Chestnut", the Tableau technique that Ada wants to share with everyone is: Ingeniously making a classification of Sangki.

Here, we will simply teach you to make a brand tree so that you can master the basics of this chart.

Specific steps are as follows:

Step 1：prepare the data sheet

The t-table has only two columns, 'connected' and 't', the connections are all links, t is from -8 to 8, with an interval of 0.5.

In the data sheet, the brand name, industry, and brand value are the original data.

In addition, the brand ID indicates the location of the brand, and the industry ID represents the location of the industry. This part of the data can be customized. Add a column name connection in order to join the t value table.

Step 2：connect data

Import the data into Tableau, select the connection field with ‘connection’, and connect internally.

Go to the worksheet and add two fields: 

Field name : logistic

Calculation:

1/(1+EXP(1)^-[t])

f(t)

[Brand ID]+(([Industry ID] - [Brand ID])*[logistic])

Step 3：making a trunk chart

Drag t to the row, f(t) is dragged into the column, and the industry, brand name is dragged to the tag card details, but the graph does not come out. Here, you need to modify the f(t) field in the column to be a dimension:

Get the following picture:

Make some adjustments, such as: drag t to the size of the tag card; set the t-axis to reverse; drag the industry into the color. Then the chart becomes like this:

Step 4：making treetop chart

Then, we will do two simple worksheets as follows:

Step 5： the combination completes the dashboard

You can get a simple brand tree by splicing in the dashboard.

Today's Tableau tips, are you getting it? Open your Tableau and try it out!