In this blog post, we will discuss the power of online visualization and a specific tool we use to that end, D3.
The power of online visualization begins with the incredible image processing power of humans. In fact, our brain can process images at millisecond speeds. While this immense visual processing capacity may have its roots in our need to survive, we believe that it offers a new potential to exponentially increase the usefulness of an ever-growing amount of online data when combined with rigorous numerical and statistical computational algorithms. Here is why we think that:
More often than not, it is unclear that the raw information from sources like connected electronic devices or social media can be of much help, or that having a machine crunch data blindly will yield anything interesting. However, the ability to display such data online so a person can interactively filter, manipulate and rearrange the data can open doors to new discoveries of trends, correlations and patterns. Having the human suggest data connections in broad strokes that can be checked by the machine can greatly shortcut the task at hand by asking relevant questions instead of all questions.
Data-Driven Documents Platform and Highlights
In our group, we have lots of – probably too much – interesting hard disk drive data to process and so we rely on online interactive visualization. We’ve found the Data-Driven Documents (D3) to be extremely useful. This JavaScript library is readily implemented in webpages and serves as a tool for online interactive representation of any dataset.
Some of the highlights of D3 are:
We believe that the powerful visualization offered by D3 in addition to computational, statistical and numerical methods to prove or disprove connections offers a new and exciting avenue that combines the best of human and machine computing.
http://gengwg.blogspot.com/
The power of online visualization begins with the incredible image processing power of humans. In fact, our brain can process images at millisecond speeds. While this immense visual processing capacity may have its roots in our need to survive, we believe that it offers a new potential to exponentially increase the usefulness of an ever-growing amount of online data when combined with rigorous numerical and statistical computational algorithms. Here is why we think that:
More often than not, it is unclear that the raw information from sources like connected electronic devices or social media can be of much help, or that having a machine crunch data blindly will yield anything interesting. However, the ability to display such data online so a person can interactively filter, manipulate and rearrange the data can open doors to new discoveries of trends, correlations and patterns. Having the human suggest data connections in broad strokes that can be checked by the machine can greatly shortcut the task at hand by asking relevant questions instead of all questions.
Data-Driven Documents Platform and Highlights
In our group, we have lots of – probably too much – interesting hard disk drive data to process and so we rely on online interactive visualization. We’ve found the Data-Driven Documents (D3) to be extremely useful. This JavaScript library is readily implemented in webpages and serves as a tool for online interactive representation of any dataset.
Some of the highlights of D3 are:
- A high degree of interactivity
- Support for different data formats, for example spreadsheets or JSON
- An ability to carry out many basic statistical calculations simultaneously with the visualization
- Great community support and lots of open source examples (for example, the D3 Gallery) that make it easy to get started without needing to learn the library in depth
- Ease of switching between many different visualizations
- No additional dependencies; other libraries can be added, removed, or changed without worrying about the effect on D3, and vice versa
- Use of SVG instead of HTML Canvas. Benefits include:
- Easier manipulation and animation, because unlike a canvas it does not need to be completely redrawn to change a single element
- SVG support for CSS, so the visuals can be adjusted without touching the JavaScript code and can easily be made consistent across the application
- SVGs are generally faster, because the browser can handle SVG rendering natively, while canvas rendering requires JavaScript to do most of the work
We believe that the powerful visualization offered by D3 in addition to computational, statistical and numerical methods to prove or disprove connections offers a new and exciting avenue that combines the best of human and machine computing.
http://gengwg.blogspot.com/
Comments
Post a Comment
https://gengwg.blogspot.com/