Skip to content

LyonDataViz/MOS5.5-Dataviz

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

90 Commits
data
data
 
 
img
img
 
 
README.md
README.md
 
 
thumbnail.png
thumbnail.png
 
 

Repository files navigation

Interactive Data Visualization (ECL MOS 5.5)

Romain Vuillemot, LIRIS, École Centrale de Lyon/Département Math-Info, Website, Twitter.

Contact or questions: [email protected]

Reading

Interactive Data Visualization for the Web
by Scott Murray

Other books

Grading

  • 30% Final exam
  • 60% Final projects
  • 10% Homeworks/Submitted assignments (pass/fail policy)
  • 10% Bonus (class participation, best homeworks/assignments)

Submissions

All homeworks/assignements/reports are due the day before the class at 23.59pm Lyon Time (GMT+1). Using this form.

Lecture 1 - Introduction to Data Visualization

Friday 10/01/2020 13:30-15:30

  • Overview: Class organization (slides)

  • Basics of Data Visualization: Perception, cognition, Visual mapping, Standard charts (slides)

  • Authoring visualizations: Libraries, Tools, Tableau Software (slides)

  • Introduction to D3.js (slides)

  • Final projects description (slides)

Tutorial 1 - Tableau Software

Friday 10/01/2020 15:45-17:45

Tableau Tutorial

  1. The goal is to have a first experience with Tableau and build standard charts using a simple dataset.
  2. Download and install Tableau Public (Free) on your machine.
  3. Altenartive for Linux users is the Online version of Tableau
  4. Other(simple) alternative to Tableau: Polestar

Problem 1: Iris flowers visualization

  1. Download the iris.csv and load it in Tableau; convert data types (if needed)
  2. Plot a scatterplot with X:sepal_length, Y:sepal_width, color:species and a trend line
  3. Save as a tab and save the workbook

Problem 2: Elections map

  1. Download the us-elections-history.csv and load it in Tableau; convert data types (if needed)
  2. Plot a grid plot with Year as columns, State as rows and State Winner as color/marks.
  3. Save as a tab
  4. Plot a geo-map with colors winning party in 2012 Latitude (generated) et Longitude (generated), with State as shapes and color ATTR([State Winner])
  5. Save as a tab and save the workbook
  6. Tips: make sure you parse the dataset correctly (FR version of Tableau automatically splits comas)

Problem 3: Stock markets visualizations

  1. Download the stocks.csv and load it in Tableau; convert data types (if needed)
  2. Plot a multiple line chart over time, for all stocks in a different color, grouped by company
  3. Plot a grouped bar chart (companies as categories, grouped by year or by companies)
  4. Your own chart!
  5. Save as a tab and save the workbook

Problem 4: Global Superstore Dataset

  1. Download the Global-Superstore-Orders-2016.xlsx and load it in Tableau; join datasets (if needed)
  2. Find an interesting story / selection with this dataset
  3. Create a Dashboard and explain your story/finding
  4. BONUS: Add storytelling (Tableau Feature)
  5. BONUS: Join other datasets (e.g. People, ..)

📅 For next class (17/01/2020)

✍ Assignments

  • PROJECT: Group proposal of 2 or 3 students

    • Submit a project topic (1-paragraph): what is the question you expect to answer? which data do you need? how do you plan to collect your data? what are the main risk in collecting/visualizing it? what are the privacy/ethical issues?

    • Create a GitHub account

📖 Readings and preparation

Tutorial 2 (1/2) - Building visualizations with D3.js

Friday 17/01/2020 13:30-15:30

var margin = {top: 20, right: 10, bottom: 20, left: 10};

var width = 960 - margin.left - margin.right,
    height = 500 - margin.top - margin.bottom;

Assignment: draw a red rectangle (like the gray one above) in the center of the page (using margin, width and height).

SOLUTION

Bar chart | SVG

  • Mark: <rect>
  • Position: x and width
  • Height: y and height
  • Color: fill (style)
  • Scales: quantitative d3.scaleLinear() and categorical d3.scaleBand()

Assignment: build a bar chart with a random dataset

  • How to generate a random list of integers: d3.range(n).map(Math.random)
  • The x.bandwidth() function generates the width attribute for the <rect>

SOLUTION

Line chart | SVG

  • Mark: <path>
  • Position: d3.line().x(function(d) { }).y(function(d) { }) to generate attribute d
  • Color: style fill
  • Interpolate: d3.line().curve(d3.curveCardinal)
  • Scales: d3.scaleLinear()

Assignment: build a line chart using a simple JSON temporal dataset add circles for each time point.

  • Create a dataset.json file in blockbuilder:

[{"id" : 1, "name": "A", "value": 10, "date": "2016-01"}, {"id" : 2, "name": "B", "value": 30, "date": "2016-02"}, {"id" : 3, "name": "C", "value": 20, "date": "2016-03"} ]

  • Be careful: one single mark (a line) to draw for the whole dataset!
  • Load a JSON dataset: d3.json("dataset.json", function(error, data) {}
  • Data parsing: d3.timeParse("%Y-%m")
  • Temporal scale d3.scaleTime()
  • Data display: d3.timeFormat("%b %y")

SOLUTION

Scatterplot | SVG

  • Mark: <circle>
  • Position: cx and cy
  • Area: r (radius) and d3.scaleSqrt()
  • Color: fill (style) and d3.scaleOrdinal(d3.schemeCategory20) color scale

Assignment: build a scatterplot using the Iris dataset and load the chart using a function that takes the chart visual mapping and dimensions as input parameters.

  • iris.csv
  • Load a CSV dataset: d3.csv('iris.csv', function(error, data){}

SOLUTION

Tutorial 2 (2/2) - Building visualizations with D3.js

Friday 17/01/2020 15:45-17:45

  • Advanced D3.js: page layout, d3.nests, legends, scatterplot and grouped bar chart (slides)

Axis | SVG

var xAxis = d3.axisBottom()
    .scale(x);

var yAxis = d3.axisLeft()
    .scale(y);

svg.append("g")
      .attr("class", "x axis")
      .attr("transform", "translate(0," + height + ")")
      .call(xAxis)

Interaction

  • Mouse click: .on("click", function(d){})
  • Mouse hover: .on("mouseover", function(d){}) and .on("mouseout", function(d) {})
  • Drag & drop d3.drag, tooltip (example)

Legends

  • Display the unique values of an attribute (e.g. use the color.domain() to retrieve them)
  • Show the visual mapping for those values (color, shape, etc.) as rows (create a group <g> and then fill the rows with the mappings)
  • You have to do it yourself!

Multiple views

  • Load and coordinate multiple visualizations
  • Charts coordination

Assignment: build a coordinated scatterplot matrix using the Iris dataset.

Submit the blockbuilder link to submit here at the end of the class (17h45)

📅 For next class (24/01/2020)

  • CODE: Submit your blockbuilder for scatterplot matrix ✉️ submission form

  • CODE: Extend the line chart using the stocks.csv where each line is a different symbol (and a different color). BONUS: add a title, legend and interactivity (e.g. tooltip, show time points, etc.) submission form

  • PROJECT: Write a document that describes the dataset(s) for your project, mechanism to collect the data and expected model (as a table, use Excel). Add more informations:

    • Project full name, project description and members names
    • Data you plan to use, does it exist, if not how you collect it?
    • 3-5 questions you want to answer using your project
    • 5-10 visualizations that are related to your project (screenshot and link sources)
    • Wait for validation by instructor before any design/coding

  • READINGS: Chapter 4. Setup, Chapter 5. Data, Chapter 6. Drawing with Data, Chapter 7. Scales.

  • What is visualization research? literacy?

Lecture 3 - Advanced D3.js and Layouts

Friday 24/01/2020 13:30-15:30

Tutorial 3: More D3.js grouping and layouts

Friday 24/01/2020 15:45-17:45

  • Grouping data with d3.nest
{symbol: "MSFT", date: Sat Jan 01 2000 00:00:00 GMT+0100 (CET), price: 39.81}
{symbol: "MSFT", date: Tue Feb 01 2000 00:00:00 GMT+0100 (CET), price: 36.35}
{symbol: "MSFT", date: Wed Mar 01 2000 00:00:00 GMT+0100 (CET), price: 43.22}

Assignment: Nest stocks by symbol and calculate aggregated values (max/min/sum) over price; parse dates.

Expected result:

0: {key: "MSFT", values: Array(123), maxPrice: 43.22, sumPrice: 3042.6}
1: {key: "AMZN", values: Array(123), maxPrice: 135.91, sumPrice: 5902.4}
2: {key: "IBM", values: Array(123), maxPrice: 130.32, sumPrice: 11225.13}

SOLUTION

Grouped bar chart |

  • Grouping: d3.nest
  • Mark: <rec>
  • Positions: using two categorical scales d3.scaleBand()
  • Color: style fill

Assignment: build a grouped bar chart using the stocks.csv .

Start using random data

var n = 10, // number of samples
    m = 5; // number of series

var data = d3.range(m).map(function() { 
  return d3.range(n).map(Math.random); 
});

SOLUTION

Stacked bar chart |

  • Grouping: d3.stack
  • Mark: <rec>
  • Positions: nested categorical scales d3.scaleBand()
  • Color: style fill
  • Scales: d3.scaleLinear()

Assignment: build a stacked bar chart using the stocks.csv .

  • Start with random data (see grouped bar chart)
  • Nest data
  • by year d.date.getFullYear()
  • by symbol
  • Calculate sum
  • Flatten the dataset to be used by the d3.stack layout
0: {MSFT: 356.07999999999987, AMZN: 527.17, IBM: 1162.97, AAPL: 260.98, year: "2000"}
1: {MSFT: 304.17, AMZN: 140.87, IBM: 1163.6200000000001, AAPL: 122.11000000000003, year: "2001"}
2: {MSFT: 261.92, AMZN: 200.68, IBM: 901.4999999999999, AAPL: 112.89999999999998, year: "2002"}
  • Apply the d3.stack() layout using the list of unique symbols as keys and the flat dataset as data

SOLUTION

Animated transitions

  • Add animation using: .transition(duration), and .delay(duration)
  • Triggered by a widget, e.g. a radio button
  • Examples of transitions: bar chart, D3 show reel).

Assignment (BONUS): build an animated transition between grouped bar chart and stacked bar chart.

  • Isolate each layout as two function grouped and stack
  • Add a swap function between each other
<div>
  <label><input type="radio" name="mode" value="grouped" checked>Grouped</label>
  <label><input type="radio" name="mode" value="stacked">Stacked</label>
</div>
  • Bind events using d3.selectAll("input").on("change", function() {})

SOLUTION

📅 For next class (31/01/2020)

  1. Write a document for your project data cleaning and preparation: data source, data shaping, processing, etc. If you use external tools (e.g. Excel, DataWrangler, Tableau) add some details of the role and steps performed using those.

  2. Load a clean data sample using d3 and descriptive charts (histogram, scatterplot, ..) in a wepage showing the characteristics of the dataset: distribution, statistics, trends, etc. Add this link to your analysis in the class document (the page should be hosted on GitHub).

  3. Draw a mockup of your project using pen and paper and add this link to the class document (the page should be hosted on GitHub).

Lecture 4 - Advanced Layout, Data Cleaning and Case studies

Friday 31/01/2020 13:30-15:30

  • Design and case studies in visualization (slides)

  • Introduction to Sketching, Rapid Prototyping, Development cycles (slides) and using the Five Design Sheet methodology

Tutorial 4: Geo-maps, design project setup

Friday 31/01/2020 15:45-17:45

  • Apply the 5DS to your project (paper and pen!)

  • Project setup using modern web development tools: local server, package managers (slides)

Geo-Map | Example

  • Mark: <path> + d3.geoPath()
  • Position: d3.geoPath()
  • Color: fill (style)

Assignment: build a geo-map following those instructions.

SOLUTION

At the end of the tutorial:

  • Update your proposal after feedback from instructor
  • Add the link to (a pdf of all) the design sheets in the class document

📅 For next class (07/02/2020)

✍ Assignments

  • Website fully functional with full dataset
    • Implement the page layouts
    • Implement graphics specific to your project

Lecture 5 - Graphs

Friday 07/02/2020 13:30-15:30

  • Graphs, Networks and Tree visualizations (slides)

Node-Link Graph | Example

  • Mark: <circle>
  • Position: d3.forceSimulation()
  • Color: fill (style)

Assignment: starting with this Node link diagram that is based on force layout with different layouts: random, radial, line, line by category and encodings: color, size.

SOLUTION

Projects (1/3)

Friday 07/02/2020 15:45-17:45

Projects (2/3)

Friday 14/02/2020 13:30-15:30

Peer-review of other groups projects

Pre-Requirements:

  • Link to description of the project
  • Link to design sheets/mockups
  • Link to a demo hosted on GitHub

Peer-review:

  • Are all the pre-requirements met? If not, what is missing?
  • Summarize the projet (what you understand): who is the audience? which question does it address?
  • Is the design effective? What would you recommend?
  • Are you aware of a similar project that would help the project? If so, please add links to it.
  • What are the strength of the project?
  • What are the weaknesses?
  • What would you have done differently?
  • Any other remark?

USE THIS FORM

Projects (3/3)

Friday 14/02/2020 15:45-17:45

You may also add the following header to your project:

<!DOCTYPE html>
<html lang="en">
<head>
    <title>YOUR_TITLE</title>
    <meta name="description" content="Your Description">
    <meta name="author" content="Authors">
    <meta property="og:title" content="Your Title">
    <meta property="og:description" content="Your Description">
    <meta property="og:image" content="https://theo-jaunet.github.io/MemoryReduction/assets/thumbnail.png">
...
</head>

In the class document:

  • PROJECT_TITLE: title of the project (10 words max)
  • PROJECT_DESCRIPTION: details of the project (2/3 sentences)
  • PROJECT_AUTHORS: list of authors
  • PROJECT_TWEET: tweet/social media-like presentation (see examples)

Projects: Final projects presentations & demos

Friday 21/03/2020 13:30-17:45

Each group has a 15min time slot (10min presentation, 5min questions) to present their project.

No slide: just show the visualization and tell a convincing story (e.g. don't list features, etc.). Should address the following:

  1. Present context, dataset, data collection
  2. Describe key design decisions (visual mappings, interactions, animations, ..)
  3. Did the visualization help you find anything of interest in the dataset?
  4. Discuss technical challenges, limits, what you would have done with more time.

Add a README.md file in the repository organized as follows:

  • Name of the participants of the project
  • Description of the project
  • Screenshots and/or video of the main features
  • Credits to external code or data you may have used in your project
  • Link to any document/report that may be related to your project

IMPORTANT -- Regarding the dataset

  • If you don't want to share your dataset, add a demo/fake dataset (e.g. sample or fake data) for the live demo
  • Provide a way to let users use their own dataset and explain how to do it (upload button? using a Python script? etc.)
  • If you take screenshots with the real dataset make sure you preserve individuals privacy

As a general rule keep in mind the projects will be made public so anybody should be able to understand on their own and privacy of the datasets should be preserved.

Projects: Polish and final submission (autonomie)

Friday 28/02/2020 13:30-15:30

Projects: Polish and final submission (autonomie)

Friday 28/03/2020 15:45-17:45

Final project is due

Exam

Wednesday 11/02/2020

  • 2h written exam
  • Questions in French
  • Answers either in French or English
  • Bring drawing material (pen, eraser, eventually colors)

Online resources

Tableau Software

D3.js

SVG

Git/GitHub

JavaScript

Data Visualization Classes

Blogs

Books

Graphics/Journals

Color

Misc

About

lyondataviz.github.io/MOS5.5-Dataviz/

Resources

Readme
Activity
Custom properties

Stars

38 stars

Watchers

16 watching

Forks

41 forks
Report repository

Releases

No releases published

Packages

No packages published