Skip to content

Tableau Cheat Sheet

Quick reference for Tableau Desktop — connecting data, building views, calculated fields, and performance.

Core Concepts

Term Meaning
Dimension Categorical field (text, date, boolean) — used for grouping/filtering
Measure Numeric field (quantitative) — used for aggregation
Continuous Blue pill — draws an axis, shows a range
Discrete Green pill — creates headers, shows distinct values
Mark A single data point (bar, point, line segment)
Level of Detail (LOD) The granularity of the view

Connecting Data

  1. Connect → choose source (Excel, Text File, Server → PostgreSQL/MySQL/SQL Server)
  2. Data Source tab: drag tables, configure joins
  3. Live vs Extract: Live = real-time queries; Extract = fast, offline copy
  4. Joins: drag table next to existing → choose join type (inner/left/right/full)
  5. Unions: drag table below existing → combine rows from similar tables
  6. Relationships (Tableau 2020.2+): flexible connections without predetermined joins

Building a View

  1. Double-click a field to add to view
  2. Drag to Rows / Columns shelf
  3. Drag a measure to the view — Tableau auto-aggregates (SUM default)
  4. Change aggregation: right-click pill → Measure → choose SUM/AVG/COUNT/etc.
  5. Change chart type: Show Me panel (top right) → click chart icon

Common drag patterns:

Result Action
Bar chart Date → Columns, Revenue → Rows
Line chart Date (continuous) → Columns, Revenue → Rows
Scatter plot One measure → Columns, another → Rows
Map Geographic field → double click (auto-places on map)
Heatmap Dimension → Columns, Dimension → Rows, Measure → Color

Filters

Type Where When to use
Data Source filter Data Source tab Filter before loading (performance)
Extract filter When creating extract Reduce extract size
Dimension filter Filters shelf Limit which categories appear
Measure filter Filters shelf Filter by value range or condition
Context filter Right-click → Add to Context Makes other filters dependent
Quick filter Show Filter on a field Give users interactive control

Filter order: Data Source → Extract → Context → Dimension → Measure → Table Calc

Calculated Fields

To create: Analysis → Create Calculated Field (or right-click in Data pane)

Basic calculations

// Order size classification
IF [Total] > 300 THEN "Large"
ELSEIF [Total] > 100 THEN "Medium"
ELSE "Small"
END

// Profit margin
([Revenue] - [Cost]) / [Revenue]

// Days since last order
DATEDIFF('day', [Order Date], TODAY())

String functions

UPPER([Customer Name])
LOWER([Email])
TRIM([Product Name])
LEFT([Order ID], 4)
CONTAINS([Email], "@gmail")
SPLIT([Full Name], " ", 1)  // first name

Date functions

YEAR([Order Date])
MONTH([Order Date])
DATETRUNC('month', [Order Date])   // first day of month
DATEADD('month', 1, [Order Date])  // add 1 month
DATEDIFF('day', [Order Date], [Delivery Date])

Table calculations (run on the view, not the data)

// Running total
RUNNING_SUM(SUM([Revenue]))

// % of total
SUM([Revenue]) / TOTAL(SUM([Revenue]))

// Rank
RANK(SUM([Revenue]))

// Previous value (like SQL LAG)
LOOKUP(SUM([Revenue]), -1)

// Window average
WINDOW_AVG(SUM([Revenue]))

Level of Detail (LOD) Expressions

LOD expressions compute at a different level from the view.

// FIXED: always compute at specified level (ignores view filters)
{ FIXED [Customer ID] : SUM([Revenue]) }   // each customer's total, on every row

// INCLUDE: add a dimension to the detail level
{ INCLUDE [Product] : AVG([Unit Price]) }

// EXCLUDE: remove a dimension from the detail level
{ EXCLUDE [Quarter] : SUM([Revenue]) }  // ignores the quarter filter

Common use: show each customer's revenue vs the overall total on the same view.

Dashboard Design

  1. New Dashboard sheet (bottom tab) → set size
  2. Drag sheets onto canvas
  3. Use Horizontal/Vertical containers for layout control
  4. Add filters: right-click a filter on a sheet → Apply to Worksheets → All using this Data Source
  5. Actions: Dashboard → Actions → add Filter/Highlight/URL action for interactivity
  6. Device layouts: Dashboard → Device Preview → customise for phone/tablet

Layout tips: - Floating vs Tiled: Tiled scales automatically; Floating gives pixel control - Use blank objects as spacers - Test with "Fit → Entire View" to ensure nothing is cut off

Performance Tips

Problem Solution
Slow dashboard Use an extract instead of live connection
Slow calculated field Push calculation to the database (custom SQL)
Too many marks (>100k) Aggregate — use Level of Detail or reduce date granularity
Slow filters Add a Context Filter for high-cardinality dimensions
Large extract Use incremental refresh, add date filter to extract

Keyboard Shortcuts (Desktop)

Shortcut Action
Ctrl + Z Undo
Ctrl + W Clear sheet
Ctrl + Alt + C Duplicate sheet
Ctrl + D Connect to data source
Ctrl + E Edit calculated field
Ctrl + Shift + B Show/hide bookmarks
Alt + drag Multi-select marks on canvas
Right-click measure → Show Me Quick chart suggestion
Ctrl + 1 Rows shelf
Ctrl + 2 Columns shelf