Create Dendrogram chart using Tableau for Data Visualization
Create Dendrogram using tableau desktop before we proceed it is important for you to know that better you are beginner or an expert you can easily follow along to build this chart with us.
A Dendrogram is a network structure it is usually created as an output from hierarchical clustering. The Dendrogram is highly used to work out the best way to assign objects to clusters it is structured in the form of a root note which for this stretch in several nodes linked by the edges or branches the last note of the hierarchy are called leaves this chart can be used in various domains like
· Health care
The Dendrogram chart broadly has two versions single level and multi-level for this demonstration we will create a single level Dendrogram visualization.
To show sales across different sub categories the data densification is required to draw the curves in the Dendrogram chart.
To perform data densification create an excel sheet with the data as shown below and name it as path.
we will begin with connecting to the sample superstore data set first this data set is provided by default when we install tableau from the sample superstore and drag the order sheet to the data source pane.
Note: If you are using tableau 2021.3 then the default data source pane will access relationships add a new connection and locate the path excel sheet for creating this chart.
We will be using joints to access joints double tap the order sheet after dragging it to the data source pane this will create a cross database join between order sheet and path sheet however as there are no common columns in the join sheet we get an error the joint has no join clauses shown below
Click add new join class select create joint calculation and Type 1 and click OK.
This joint densifies your data that is used to define the hierarchical view in the Dendrogram chart this will double the number of rows
Let’s go to Sheet 1 right clicks on path measure select create bins set the bin size as 1 provides a suitable name for the bin, we will use the default name here that is path bin.
Next we will create the following calculated fields required to create the Dendrogram chart name the calculated field as sales value enter the calculated field as shown below.
Window sum is used to calculate the sales value for the entire view ask the data is densified the number of rows gets stumbled so to get the actual values we are dividing the whole value by 2.
Name the next calculated field as rank enter the calculated field as shown on the below this will assign unique rank to the sales value in ascending order.
we will create another calculated field and name it as it is X the expression for X is as shown on the below this expression will be calculated for each of the point based on the index from minus 6 to 18 with a step size of 0.12 this is required for the sigmoid function that we will use to create a curve line 0.12 also signifies the diverging point for each curve.
Name the next calculator field as sigmoid expression for this calculated field is a show on the below this is the sigmoid function that will take the X points and plot the exponential curves.
Final calculated field by expression for Y is as shown on the below this calculation uses the sigmoid function created previously to give the desired height for each curve after the calculations.
Let us start creating a data visualization change the mark type to line drag Path bin onto the column shelf and right click on the bin and ensure that so missing values is selected and drag it onto the path shelf drag X onto the column shelf and right click on the pill go to compute using select path Bin will help us get the proper curves for each path.
values that we have drag Y on to the rows shelf right click on this pill go to compute using select path Bin then drag sub category to label mark. let us know said the granularity of rank and Y to sub category to achieve the height of each curve and rank each sub category on the basis of sales to do so right click on the white pill and select edit table calculation in nested calculations select rank in compute using specific dimensions and select only sub category as shown in below
Now in nested calculation select Y in compute using select specific dimensions. Select only sub category.
we had initially said the granularity level of Y to compute at path Bin hence all the four measures that are XY rank and sales value are being computed at path Bin leveled however to plot the proper curves we have said the Y and rank to compute at sub category level.
And drag sales value onto the color mark and right Click to select compute using and select path Bin next drag sales value to size mark.
This will set the size of each Curve on the basis of sales and select entire cosmetic changes as shown on the below to achieve the final result.
Follow us for more such topics and suggestions.,
Also click the link below to follow us in LinkedIn https://www.linkedin.com/showcase/sphyzee-analytics/about/