What is this ?
This project enables people to chart data and publish it.
The idea is to popularize statistical information using data vizualisation. It starts from the observation that:
- statistical facts are published by institutes for statistics, survey firms or the open data stream; but this statistical data is either published as complex reports only read by some users, or provided in a raw form (as for open data[1]). This data is not sexy enough to be read.
- many people are interested by statistical facts as they help us to understand the world they live in - or because this information is simply fun.
- technical tools are available to create nice charts, like the d3js library. But these tools require some programming abilities not shared by everybody.
In this context, we could expect available data to be displayed in an attractive way using the powerfull tools; but to date, only some newspapers are publishing them.
So in practice this project proposes to chart data from independant studies and publish these charts.
[1] http://en.wikipedia.org/wiki/Open_data
Philosophy
- attract the audience:
- do nice charts which provide a feeling of the message, not only some numbers.
- use an accessible tone in description; irony and jokes are part of what makes the content fun; balance this familiar tone with a serious and sourced references.
- don't lie:
- always provide references to guarantee the information is somewhat correct, and to enable readers to understand where this data comes from, how it was collected, by who and for who.
- publish no original work, only studies published by independant third parties.
- When different studies lead to different results, mention them all.
- When a study is debated, mention and link this debate (or don't publish anything based on this study ?)
- inform:
- don't only provide raw statistics, but also provide explanations for a wide audience to understand the topic, the vocabulary, the implications of the topic.
- Provide curious readers with links to go further
- Don't only reproduce only one simple statistic; also add some context, compare it with other results.
- opinions are not facts: this is not a place for politic debate, propagation of ideology.
Original chart: silhouettes
We provide a chart which is not basically proposed by third-party libraries: the "1over100" chart.
Terms of use
For personal use only.
This website is provided as it, without any guarantee of availability.
Inspiration
Data vizualisation
- Sankey diagrams: http://bost.ocks.org/mike/sankey/
- Four Ways to Slice Obama’s 2013 Budget Proposal: http://www.nytimes.com/interactive/2012/02/13/us/politics/2013-budget-proposal-graphic.html?_r=0
Roadmap
This project is in alpha stage (work in progress).
Next steps include:
- provide a GUI for people to input data themselves
- find an ergonomic way to search and/or filter data
- find a way to facilitate the edition of existing charts (so, version charts)
- add novel chart types
- think about the collaborative aspect: what is required ? Workflow ?
- find someone able to deal with webdesign to make the site beautiful
- create index pages with only data in certain languages to be displayed
- prepare the website for referencing (opengraph)
Past steps:
- technical feasability display a simple chart
- provide several examples of charts to illustrate what the spirit of the tool is
- prototype the publication technics to chart stuff
- make several persons interested in reading charts and adding some
technical aspects
philosophy
- be quick
- be simple: single-page layout
- most load is reported in the client side: rendering of charts, user forms
- one-page layout: the first display of the webside loads data
- cutting-edge technology: data is rendered using the last technologies (HTML5, SVG, etc.)
tools
This website relies on many tools.
About formats, the website uses:
- HTML5
- CSS 2
- javascript
- JSON
- mysql
- apache2
Friends
... or people we would like to be friends with.
XKCD is a comics which frequently includes nice charts : http://xkcd.org/
existing libraries
Here is a state of the art of the librarys which exist to date:
already used:
candidate for use:
- google charts: https://developers.google.com/chart/interactive/docs/gallery?hl=fr
- chartjs: http://www.chartjs.org/
- dojox: http://livedocs.dojotoolkit.org/dojox/charting
- prefuse flare: http://flare.prefuse.org/
- flot: http://www.flotcharts.org/
- rgraph: http://www.rgraph.net/
- zingchart: http://zingchart.com/gallery/
- flotr2: http://humblesoftware.com/flotr2/
- cubismjs : http://square.github.io/cubism/
- xkcdgraphs: http://dan.iel.fm/xkcd/ http://xkcdgraphs.com/
- signma.js http://sigmajs.org/
- datamaps http://datamaps.github.io/
- snap svg http://snapsvg.io/
other libraries which are not candidate for use:
- Raphael (no more maintained) and grephael: http://raphaeljs.com/ http://g.raphaeljs.com/
- Highcharts (non free): http://www.highcharts.com/
- protovis is now replaced by d3js http://mbostock.github.io/protovis/
- olapchart (non free) http://www.olapcharts.com/
sources:
- http://stackoverflow.com/questions/119969/javascript-chart-library
- http://en.wikipedia.org/wiki/Comparison_of_JavaScript_charting_frameworks