Empowering Business Intelligence: The Tableau vs#

3.1 Installation#

You can download Tableau Desktop from the official Tableau website. Depending on your operating system (Windows or macOS), Tableau’s setup wizard guides you through the standard installation procedure.

3.2 Interface Overview#

Upon launching Tableau Desktop, you’ll see several main components:

• Data Pane: On the left, you will find the data pane where your connected data sources and fields are listed (dimensions and measures).
• Shelves: Rows and columns shelves for placing fields.
• Marks Card: Controls color, size, text, detail, and tooltips of data points.
• Canvas/Worksheet: The central area where you build and visualize your data.

3.3 Data Terminology#

• Dimensions: Typically categorical data (e.g., customer name, city, product type).
• Measures: Typically numeric data (e.g., sales amount, profit, quantity).

Example: Connecting to Excel Data#

Let’s assume you have an Excel file named SuperstoreSales.xlsx containing sales orders data. Below is a sample structure of the file:

To connect:

1. Open Tableau Desktop.
2. Under the “Connect” pane, select “Microsoft Excel.”
3. Browse to your file, SuperstoreSales.xlsx, and open it.
4. Drag the relevant sheet into the data source canvas.

Once your data is on the canvas, you can perform joins or unions if required. Click “Sheet 1” at the bottom to start building your first visualization.

5.1 Basic Steps#

1. Open a Sheet: After connecting your data, go to a new sheet.
2. Choose Fields: Drag a dimension (e.g., Region) onto the “Columns” shelf and a measure (e.g., Sales Amount) onto the “Rows” shelf.
3. Change Visualization Type: To choose between bar, line, scatter, etc., use the “Show Me” panel or select from the Marks card.
4. Add Interactivity: Drag another dimension (e.g., Product Category) to color in the Marks card. This visually separates sales per category.

5.2 Example Visualization: Sales by Region#

• Drag Region to Columns
• Drag Sales Amount to Rows
• Drag Product Category to Color
• Set Marks to “Bar”

This will produce a stacked bar chart showing sales by product category across different regions.

6.1 Creating a Dual-Axis Chart: Profit vs Sales#

Drag Sales Amount to Rows, drag Profit to Rows again (or right-click and choose “Dual-Axis”). Synchronize the axes if needed. Configure lines or bars to suit your best representation. For instance:

• Measures: Put Sales Amount as bar, Profit as line.
• Color: Differentiate them using color in the Marks card.

7.1 Basic Calculations#

Tableau supports simple drag-and-drop calculations, such as:

• Aggregations: SUM, AVG, COUNT, etc.
• Logical Functions: IF, ELSEIF
• String Functions: CONCAT, LEFT, RIGHT

A common calculation might be computing profit ratio:

1
[Profit Ratio] = [Profit] / [Sales Amount]

Use “Create Calculated Field” in the Analysis menu or by right-clicking in the data pane.

7.2 Table Calculations#

Table calculations are performed on the results of existing measures. Examples include:

• Percent of Total
• Running Total
• Moving Averages

7.3 LOD Expressions (Level of Detail)#

Level of Detail expressions allow you to compute values at different levels of the data. Examples:

• Fixed: { FIXED [Region] : SUM([Sales Amount]) }
• Include: { INCLUDE [Category] : SUM([Profit]) }
• Exclude: { EXCLUDE [Category] : AVG([Quantity]) }

These advanced calculations help you answer deeper questions. For instance, if you want to compute each region’s share of total sales within the entire dataset but then display that figure by category, you can use LOD expressions to fix the region’s total before further aggregations.

8.1 Building a Dashboard#

A dashboard is a collection of multiple worksheets in a single layout. To create a dashboard:

1. Click “New Dashboard” at the bottom of Tableau.
2. Drag Worksheets from the left pane onto the canvas.
3. Adjust sizes, add filters, legends, or other relevant elements.

8.2 Interactive Elements#

• Filters enable viewers to slice data by specific criteria.
• Action Filters let viewers click a mark in one sheet and automatically filter the other sheets.
• URL Actions can open external webpages or link to other dashboards.

8.3 Story Points#

A story in Tableau is a sequence of worksheets/dashboards that guide viewers through a narrative. This can be especially powerful for presentations:

1. Click “New Story” at the bottom.
2. Drag views or dashboards to individual story points.
3. Add captions and text to give context or highlight changes over time or among categories.

9.1 TabPy Example#

1. Install TabPy:

   1
   pip install tabpy
   2
   tabpy
   

   This starts the TabPy server on your local machine (default: localhost:9004).

2. Connect Tableau to TabPy:

   • In Tableau, go to “Help” → “Settings and Performance” → “Manage Analytics Extension Connection.”
   • Set the server to localhost and port to 9004.

3. Use Python Scripts in Calculations: In Tableau, create a calculated field using SCRIPT_XXX functions. For example:

   1
   SCRIPT_REAL(
   2
     "import math
   3
      return [math.sqrt(x) for x in _arg1]",
   4
      SUM([Sales Amount])
   5
   )
   

   This calculation sends the measure “Sales Amount” to Python, calculates the square root in Python, and returns the result to Tableau.

By incorporating Python scripts, you can tap into advanced math libraries (NumPy, pandas, scikit-learn, etc.) for machine learning or advanced statistical modeling, returning results directly into Tableau for visualization.

10.1 Tableau Server and Tableau Online#

For broader access across an organization, users typically publish dashboards to Tableau Server (on-premises) or Tableau Online (hosted in the cloud). Key features include:

• Centralized Management: Administer data sources, user groups, and permissions.
• Alerts and Subscriptions: Automatic email notifications when certain thresholds are reached.
• Browser-based Editing: Light editing and collaboration via a web browser.

10.2 Tableau Public#

Tableau Public is the free platform to share visualizations with the world. However, be mindful that all data published to Tableau Public is publicly available.

Observations#

1. Power BI often aligns well with Microsoft-driven organizations and offers a lower price point for smaller teams.
2. Qlik Sense is known for its powerful associative engine but might require more technical learning upfront.
3. Tableau thrives in pure analytics tasks, data storytelling, and is often praised for interactive visuals.

Each tool has its strengths; the best choice depends on organizational structure, budget, data complexity, and existing technology stack.

13.1 Server Administration Customization#

• Security Models: Set up row-level security to ensure that certain users only see relevant slices of data.
• Performance Tuning: Monitor Tableau Server resource usage and optimize workbook load times.

13.2 Embedded Analytics#

• Web Embedding: Host Tableau visualizations within your company’s internal portals or client-facing websites.
• JavaScript API: Control embedded Tableau objects, respond to interactivity events, and inject custom filtering from external web apps.

13.3 Data Prep with Tableau Prep#

• Data Flows: Create repeatable flows to clean, combine, and organize data before analysis.
• Scheduling: Automate your data preparation tasks using Tableau Prep Conductor (available with Tableau Data Management Add-On).

13.4 Hyper API#

Tableau’s Hyper Engine is high-performance in-memory data processing. You can programmatically create, read, update, or delete data in .hyper files via the Hyper API. For instance, a Python snippet:

1
from tableauhyperapi import HyperProcess, Connection, CreateMode, TableDefinition, \
2
    SqlType, Telemetry, Inserter, \
3
    TableName
4

5
with HyperProcess(Telemetry.SEND_USAGE_DATA_TO_TABLEAU) as hyper:
6
    with Connection(endpoint=hyper.endpoint,
7
                    database='sales.hyper',
8
                    create_mode=CreateMode.CREATE_AND_REPLACE) as conn:
9

10
        # Define table structure
11
        sales_table = TableDefinition(
12
            table_name=TableName("Extract", "SalesData"),
13
            columns=[
14
                TableDefinition.Column("OrderID", SqlType.text()),
15
                TableDefinition.Column("Category", SqlType.text()),
16
                TableDefinition.Column("SalesAmount", SqlType.double()),
17
                TableDefinition.Column("Quantity", SqlType.int()),
18
                TableDefinition.Column("Region", SqlType.text()),
19
            ]
20
        )
21

22
        conn.create_table(sales_table)
23

24
        # Insert sample data
25
        row_data = [
26
            ["1", "Office Supplies", 250.0, 5, "East"],
27
            ["2", "Furniture", 700.0, 2, "West"],
28
            ["3", "Technology", 1200.0, 3, "East"]
29
        ]
30

31
        with Inserter(conn, sales_table) as inserter:
32
            for row in row_data:
33
                inserter.add_row(row)
34
            inserter.execute()

This code programmatically creates a Hyper database file, defines a table structure, and inserts sample data. You can then connect Tableau Desktop or Server to the generated .hyper file for near real-time analytics.