Data Visualization with Tableau- Part 2

Continuing with the previous part of the series, I will continue with the other visualizations that can be made with Tableau.

Line Chart - A line chart can be used to show a trend or a correlation between the two variables in the dataset.

To draw a line chart follow the given steps,

1. Drag the 'Total' from data pane to the rows shelf on the top.

2. Drag the 'Date' from data pane to the columns shelf on the top. There change the date type to month.

3. Drag the 'Customer Type' to the colors.

This analysis shows that there is a considerable dip in the revenue and possibly sales in the month of February irrespective of whether we consider member or non-member customers.

Bubble Chart - These charts display data as circles shown on a two-dimensional chart. Each bubble represents a single data point, with its position on the chart determined by its values on two numeric axes. The bubble size can encode additional details about the data, such as the particular product sold.

To make a bubble chart follow the given steps,

1. Drag and drop 'Invoice ID' from data pane to the columns shelf. Right click and change it to the count.

2. Drag and drop 'Product Line' from the data pane to the rows shelf.

3. Choose the bubble type chart from the 'show me' card that displays the type of visualization in the Tableau on the right most top corner.

4. Drag and drop 'Product Line' from the data pane to the colors card under the marks.

5. Drag and drop 'Invoice ID' from the data pane to the text card under marks and convert it to count.

This analysis shows that the orders are more or less equally distributed across the different product categories. Relatively products in the health and beauty line have lowest sales while the fashion accessories have the highest sales.

In the above chart, it is important to use count of 'Invoice ID' to quantify total number of orders specific to each product line. If we omit this step we will still be able to get bubble charts but the size of the bubbles will not represent the total number of orders in a particular category. Instead, each bubble will show an individual order and its category as shown below,

This chart looks really pretty compared to the previous bubble chart. However, it does not convey interesting details about our data. So it is important to note that besides having a beautiful visual, it is necessary for the visual to convey the message or information about the data clearly.

Treemap - A treemap helps to visualize and display hierarchical data using nested rectangles. The area of each rectangle is proportional to the quantity it represents. The rectangles are non-overlapping so they can be clearly inferred. They aid in comparing the proportions and patterns in the data.

To create a treemap use the following steps,

1. Drag and drop the 'City' and 'Product Line' from data pane in the columns shelf to the top.

2. Drag and drop the 'Invoice ID' from data pane in the rows shelf to the top. Convert the invoice id to count

3. Choose the treemap type chart from the 'show me' card that displays the type of visualization in the Tableau on the right most top corner.

4. Drag and drop 'City' , count (Invoice ID), and 'Product Line' to the text card under the marks.

This analysis shows that in Yangon highest number of order fall in home and lifestyle category while in Naypyitaw the highest number of orders fall in food and beverages category.

Heatmap - A heatmap uses color coded table or matrix which helps to compare relative values of data points within dataset and identify patterns hidden in the data.

To make heatmap, follow these steps,

1. Drag and drop the 'Product Line' from data pane in the columns shelf to the top.

2. Drag and drop the 'City' from data pane in the rows shelf to the top. Convert the invoice id to count

3. Choose the heatmap type chart from the 'show me' card that displays the type of visualization in the Tableau on the right most top corner.

From this picture by looking at the colors in the heatmap we can clearly see that in Mandalay city higher number of orders fall in the fashion accessories and sports and travel category. In the city Naypyitaw more orders fall in the fashion accessories and food and beverages category. In contrast, higher number of order made in Yangon fall in home and lifestyle category of product line.

Text Tables - The text table provides a means to display both the non-numerical as well as numerical data in the form of a text by declaring them as dimensions. It is an easily way to display some of the information in data in written form without having to dig into a report to get numbers. The most common form of text table is table with columns and rows having data displayed as text.

To create a text table follow the given steps,

1. Drag the 'Invoice ID', 'Cutomer Type', 'Gender', and 'Product Line' from data pane into the rows shelf on the top.

This way we can present the desired information in a tabular form. The data can be further limited using the filters in Tableau. For instance, we can put a filter for the customer type and see the product line category for either the member and non-member customers. It maybe able to reflect if the non-members visit the store to buy more specific product categories.

Nevertheless, we can also add measures to the text tables to include other information. Add the measures to the column shelf. The numerical measures can be displayed as marks in the textual table and the marks can be chosen from various shapes including bars, squares, circles, and gantt bars. An example is seen in the figure below where the gross income and quantity related to each order id are included in the textual table by adding them as measures to the column shelf on the top.

As seen in the table different marks (squares and bars) have been chosen to represent the measures quantity and gross income.

Highlight Table - It is a table in which the data values are displayed differently using varied colors, fonts or symbols in a way that it becomes easy to identify trends or patterns in data just by looking. It is a valuable tool for presenting data in a clear and concise way.

To create a highlight table follow the given steps,

1. Drag the fields 'Branch', 'City', and 'Invoice ID' from data pane to rows shelf on the top. These will be dimensions.

2. Drag 'Taxes' from data pane and add it to the columns shelf. This will lead to bars with lengths proportional to the total amount for each order id.

3. Drag the 'Taxes' from data pane to the colors card under marks. This will cause the bars in the table to change colors that are representative of the numerical value of 'Taxes'.

4. Next tight click on the horizontal axis for 'Taxes' bars and edit the axis. Change the range from automatic to custom and define the range as 0 to 1. This will give bars of equal length but colored that is representative of the value of 'Taxes'

From this visual it can be inferred that the order id 139-32-4183 had the highest amount of tax paid for it as the tax bar corresponding to it is red in color. And the color code for highlight on the right shows that red color is associated with highest values of tax.

From the se discussion on the various charts, graphs and tables, we see that different visualization can be used to gain a number of interesting insights from a given dataset in Tableau with simple clicks, drags and drops. Thus, Tableau is a simple data analysis and visualization tool that can be used to draw meaning inferences from datasets.