[4]Tableau[Join|Blend|Dual Axis Charts]

Hyejin Jeon
Media Analyst : Hyejin Jeon
4 min readMay 15, 2020

1) Types of joins

Inner Join / Left Outer Join / Right outer Join / Full Outer Join

2) Join with duplicate values

3) Joining on multiple fields

4) Joining data vs Blending data in tableau

Go to this website: Superdatascience.com/pages/tableau

Using this dataset: P1-AmazingMartEU2

Table(A+B) + Table(b) : how to join in this case?

We don’t have to consider how to join these three tables! We have “blend”

First, join with “order and breakdown” two tables, and then open the same data but designated “sales” table. This is blending.

If the data source is different and you can’t join the table properly, you can use “blend” function

1) different level of granularity 2) the data type is just different data sources

5) Data blending in Tableau

Go to this website: Superdatascience.com/pages/tableau

Using this data set: Airline2 (P1-Airline-Comparison)

Blue is the primary table. Blend function is the same function of a left join

You can designate the primary table as well. Just change the drop sequence

6) Blend chart shows separately and we can control together or separately: this chart like left join on the Region

7) Label each Airline Revenue

Alert!! Marks are separately by each airline table. Thus, when you change the Revenue (FY2015) / (FY2016), you should change the Marks tab first!

When you see the right side of the chart, you can check the data is duplicated. Why?? Because tableau doesn’t know how to deal with empty data. So let’s fix this problem

8) Fix the duplicate problem: There are two ways

First, go to edit blend, Second, change the dimensions name

Data > Edit Blend Relationships > add > period > year

Then you can check the dimensions, Both Year & Region are blended

Another way to fix

Dimension > year > Period > tableau automatically changes the format

9) Dual-axis chart

Blend default: month + year: because we checked “the month of order column” included month/year

10) Build Dual axis

Right click on target > Dual axis > synchronize axis

Alert!! After clicking the dual-axis, You should do synchronize axis: range adjusts

You can get the insights from each department. If you change the orders of Rows, and the Dual-axis visual change as well. For example, when you change the spot sum(Target), sum(Sales) the sequence of the chart is also changed

11) Creating Calculated Fields in a blend

Actual sale vs Target Sale > Calculate the Excess of Target > SUM(Sales) * SUM(target> drag and drop the column)

We can check the time series or how many times revenue was under average

11) Section Recap

· Blending data

· A blend is like a smart join ‘on the fly’

· It is a left join

· Common fields with the same name are picked up automatically as the blend clause

· Blending occurs at the granularity of your view

· Aggregation happens before the blend

· Blends are unique to each worksheet

13) Joining vs Blending

Use Joins when

· combining data at the row-level

Use Blends when:

· Datasources have different levels of granularity

· Datasources come from a different system (SQL or excel)

14) Dual Axis Charts

· Remember to Synchronize the Axes

· Move charts backward and forwards as required

· Calculated fields in a Blend

· Values are aggregated due to the nature of Blends

--

--

Hyejin Jeon
Media Analyst : Hyejin Jeon

Data analytics cheat sheet | python | Tableau | Machine learning | Media data analyst