Tableau's advanced features take your data visualization skills to the next level. From complex calculations to geospatial analysis, these tools empower you to create insightful, interactive dashboards that tell compelling stories with your data.
Mastering these techniques allows you to optimize performance, design user-friendly interfaces, and uncover deeper insights. By applying best practices in dashboard design and leveraging Tableau's powerful capabilities, you can create impactful visualizations that drive informed decision-making in business intelligence.
Advanced Calculations in Tableau
Calculated Fields and Table Calculations
Top images from around the web for Calculated Fields and Table Calculations
sql - I am trying to create a calculated field from multiples columns by each row in Tableau ... View original
Is this image relevant?
sql - I am trying to create a calculated field from multiples columns by each row in Tableau ... View original
Is this image relevant?
1 of 1
Top images from around the web for Calculated Fields and Table Calculations
sql - I am trying to create a calculated field from multiples columns by each row in Tableau ... View original
Is this image relevant?
sql - I am trying to create a calculated field from multiples columns by each row in Tableau ... View original
Is this image relevant?
1 of 1
Tableau provides a robust calculation language called "calculated fields" that allows users to create custom formulas and complex logic beyond the standard aggregations and dimensions
Calculated fields can reference other calculated fields, allowing for layered logic and modularity in calculations which enables breaking down complex formulas into smaller, reusable components
Table calculations are a special type of calculated field that perform calculations across the values in the entire table, such as running totals, percent of total, rank, percentiles, moving averages, and lead/lag analysis
Calculated fields can be used to create ad-hoc groups, sets, bins, and parameters which enable users to dynamically interact with the view based on their selections (filters, highlighters)
Level of Detail Expressions and Advanced Functions
Level of Detail (LOD) expressions allow for calculating values at a different level of granularity than the view, such as calculating a total sales value for each region to use in a calculation for each sub-category within that region
The three types of LOD expressions are FIXED (compute at specified dimensions independent of view), INCLUDE (add dimensions to view level), and EXCLUDE (remove dimensions from view level)
Advanced date functions can be used to calculate different levels of date granularity (day, week, month), fiscal calendars, time series analysis, cohort analysis, and custom date ranges
Logical functions can be used to create conditional calculated fields with if-then-else logic, case statements, and boolean comparisons
Mathematical functions enable advanced statistical analysis (mean, median, standard deviation) and number manipulation (absolute value, rounding, exponential)
Data Extracts for Performance
Creating and Filtering Extracts
Tableau data extracts are compressed, columnar stores of data that are optimized for aggregation to enable faster performance, rather than querying the entire raw dataset
Data extracts can be filtered to only include a relevant subset of data for analysis, reducing the amount of data stored and time required to render views
Filters can be added to the data source page, an extract filter can be defined while creating the extract, or a extract can be filtered to include only the data in the current view
Aggregation, such as averages, counts, or distinct counts, can be defined while creating an extract to pre-summarize the data, resulting in much smaller extract sizes
Optimizing and Refreshing Extracts
Extracts can be set to incrementally refresh, only adding new rows or updating values that have changed since the last refresh, rather than processing the entire dataset each time
Tableau can automatically determine what fields to index to optimize queries or developers can manually define fields as indexes to speed up extract creation and query times
In Tableau Server, schedules can be defined to automatically refresh extracts and maintain an up-to-date copy of the data with email alerts configured if a refresh fails
Workbooks using extracts should be published with the packaged data option to improve load times and enable offline access for end users
Geospatial Analysis with Tableau
Geographic Roles and Custom Geocoding
Tableau automatically recognizes geographic fields, such as country, state, city, or zip code, in a data source and geocodes them to latitude and longitude coordinates to plot them on a map
Custom geocoding can be defined by creating a Tableau data extract from an Excel file, CSV, or text file that contains the location names and corresponding latitude and longitude coordinates
Radial selection allows for identifying data points within a specified distance of a central point, such as finding all customers within 50 miles of a store
Advanced Mapping Techniques
Filled maps encode data assigned to a geographic area by filling the area with color, based on a measure which allows for quickly identifying trends across regions (population density by state)
Symbol maps compare data assigned to specific locations by placing a marker over each location, with the color and/or size of each marker based on a measure (sales per store)
Density maps visualize concentrations within a geographic area by grouping together areas with many data points and coloring them based on a measure, such as a heatmap (crime incidents in a city)
Spider maps and origin-destination maps show movement between two locations with lines or paths, allowing for analysis of spatial patterns (flight routes, supply chain paths)
Dual axis maps allow for plotting two measures on the same map view, such as color encoding one measure and size encoding another on the same set of data points
Dashboard Design and Optimization
Visual Best Practices
Dashboards should follow a clear visual hierarchy, with the most important information and key takeaways placed prominently at the top and left, since most cultures read from top to bottom and left to right
Gestalt principles of visual perception, such as proximity, similarity, and enclosure, can be used to logically group related dashboard elements and separate distinct sections
Objects should be logically laid out in a grid with consistent padding between items using Tableau's layout containers to group items and ensure consistent spacing
Fonts should be legible with high contrast against the background and a consistent typographic hierarchy (headers, subheaders, body text) to improve readability
Interactivity and User Experience
Interactivity, such as filters, parameters, and actions, should be included to enable end users to ask and answer their own questions, but too much interactivity can be overwhelming and confusing
Guided analytics can be implemented to direct the end user through the dashboard in an author-driven flow, using a combination of interactivity and static explanations
Tooltips can be customized to show metadata, additional context, or even entirely new charts when hovering over a data point which saves space on the dashboard while still providing granular details
Accessibility features like keyboard navigation, alt text for images, and colorblind-friendly palettes ensure the dashboard can be consumed by all users
Performance Optimization
Dashboard load times can be improved by reducing the number of views and data sources, filtering the data, using Tableau data extracts, and simplifying calculations
For optimal performance, it is best practice to minimize the number of data sources for a dashboard, ideally having all data consolidated in a single data source
Limit the use of high-density mark types like high granularity area charts and maps with many data points which take longer to render
Disable automatic updates and set the dashboard to a fixed size to prevent it from re-rendering if the browser window is resized