Wilson Mar bio photo

Wilson Mar

Hello. Hire me!

Email me Calendar Skype call 310 320-7878

LinkedIn Twitter Gitter Google+ Youtube

Github Stackoverflow Pinterest

See all 4 dimensions (time) in Tableau visualizations


Overview

Understanding the trend of numbers over time is a fundamental skill. That’s why it’s taught in elementary school.

“Those who don’t know history are doomed to repeat it.” – Edmund Burke

Excellence defined

Below are charts that posses traits of excellent visualizations:

  • Dynamic data - update visualizations (in “live mode”) as data changes in sources such as databases.

  • Visual querying - change the query by selecting or clicking on a portion of the graph or chart (to drill down, for example).

  • Linked multi-dimensional visualization - selections made in one chart are reflected as you navigate into other charts.

  • Animation -

PROTIP: Dynamic (movie mode) is available only within the Tableau Public client, not when viewed on websites (as of 2016-01-05).

  • Personalization - give power users an in-depth view and newbies a simpler view, and also control access to data based on user- and role-based access privileges.

  • Actionable alerts - thresholds and parameters that trigger messages whether you’re interacting with reports or not.

Fortune 500 in the US

http://beta.fortune.com/fortune500/visualizations
dataviz time-series 2030x984

Select an industry and mouse over a line to see each company’s ranking change over the last 20 years. The higher each line appears, the higher the company is on Fortune magazine’s 500 largest public companies.

Companies that have a rising trajectory include:

  • Whole Foods Market
  • Amazon
  • Alphabet (Google)
  • Apple
  • CVS Health
  • Comcast
  • Dollar General
  • Gilead Sciences
  • Qualcomm
  • Facebook
  • Visa
  • Master Card
  • Cognizant
  • Oracle

http://beta.fortune.com/global500/visualizations

Illustrates the fast growth across industries in China.

Plus:

  • Softbank

Examining the HTML shows use of the Chartbeat JavaScript library for visualization.

Nicolas Rapp also created for Fortune 500 this graphic:

dataviz-fortune500-oil-1120x772

Hans Roling

Since 2007 at http://www.gapminder.org/videos/ Hans Roling, a professor of health statistics in Sweden, is an internet legend for his “Joy of Stats” video shown on BBC Nov 10, 2010 and Ted Talks. In them he shows his Gapminder web app which presents multiple dimensions dynamically over time (nearly 300).

Jeffrey Shaffer (@HighVizAbility, co-author of bigbookofdashboards.com) created a time lapse video of his viz showing a trail of dots which grow in size and get darker over time (as the legend notes):

tableau time lapse gapminder

Here, different colors represent different countries, with the United States in red. Most other countries saw a decrease in fertility rate over time while life expectantcy increased.

Within the wonderful Tableau gallery Andy Cotgreave (@acotgreave, now Technical Evangelist at Tableau) built over the years is this re-creation of Gapminder:

tableau gapminder

Andy explained in 2010 how he created the above using Tableau Trendalyzer v6. Download his viz from the Tableau Public website to manipulate using Tableau Public client installed on your laptop.

The trails is a recreation of (@moritz_stefaner) “Remixing Rosling” in @tableau.

Cycle Time

tableau interactive wait times 1956x1372
Click for full pop-up

Metrics include:

  • Percent of value-add time vs. total time. For example: In the case of physician visits, time interacting with physician vs. waiting and other activities.

Time Series databases

Introduction to Time Series

  • KDB proprietary runs in-memory within a single machine.

  • Graphite - legacy invented by Expedia - does automatic roll-ups of data (losing irregular values)

  • RiakTS from Basho is built for scale

  • Promethius for DevOps (highly available with scaleout?)

  • OpenTSDB is a layer on top of HBase.

  • TimescaleDB, an open source time-series database engineered from PostgreSQL,

  • InfluxDB, open-source (MIT)

    • No external dependencies (written in Go)
    • SQL-like query language
    • Input data “Line Format” (not JSON)
    • Stores data in compressed format
    • Horizontally scaleable (across several servers)

    • Kapacitor collects anomalies

See video “How to Manage TensorFlow with InfluxData” (The TensorFlow Jupyter notebook for weather prediction is shown from 17:24)

More on front-end styling

This is one of several topics:

  1. Text Editors
  2. Markdown text for GitHub from HTML
  3. 508 Accessibility

  4. JAM Stack Website Project Plan
  5. HTTP/2 Transition Project Plan
  6. Static websites
  7. Jekyll Site Development

  8. Website styles
  9. Website Styling

  10. Email from website
  11. Search within Hyde format Jekyll websites
  12. Windows Tile Pin Picture to Website Feed

  13. Data Visualization using Tableau