Data mapping

DRAFT in progress

Data > Information > Knowledge - Colleen is not about this, it's about exploring the raw, exploring the feeling/essence… the depth in data
the key is to elucidate that there is constant human effort happening to analyse the data sets for science
less visualising data, more visualising human effort

Definition of variables which require mapping to data.

Each project community/group has a theme which determines its look and feel, and a mode which defines its behaviour.

Each project has a theme and a mode for individuals within it (polygons) and for the community as a whole.

Config in client/js/chart.js

For interval set the number of seconds, on line 15. Currently its 86400 for a day, set to 3600 for an hour this.intervalSecs = 86400; // 3600; //

For date range change lines 73 and 74. The second figure for months is 0 indexed, so 5 = June and 6 = July.

this.pickerStartDate = new Date(Date.UTC(2014, 5, 10));

this.pickerEndDate = new Date(Date.UTC(2014, 6, 10));

How are the polygons placed (location)?
How do we keep the animation sparse?
Should projects appear at the same time or individually? If individually, how is the active project chosen?
What happens when there is no user activity?
When a project moves to the foreground, how does the transition work? does it scale up from its background position to the centre, and the previous foreground one scale down back to its original position? or perhaps it fades out and back in again larger?
Does everyone see roughly the same thing, i.e. does the server-side know which project is foregrounded, and updates the client accordingly?
Is there a difference in the animation rules depending on whether the project is in the foreground? NO
How are the projects positioned in the background, and should this be determined server-side? This is done with the graph network layout tool. Project relationships are defined in a CSV based on similarity of content (JF defined)

n-body choreography gives us objects orbiting in a set pattern.
-- how can we disrupt without the object falling out of the world? We can't they are fragile systems and we lose the objects if trajectory disrupted.
-- what elements of the data provide the mass and radius of the objects? The classification timeseries data can be used to control the speed of the objects
Objects can appear or disappear (in priority - ID - order) related to numbers of users on the project.
Open source JS code avail for both the below (and stack more out there):

Each project is tagged
Tag/relationship counts create a web-like structure of the projects
As a test I used this table to create a network in Fusion Tables, simply subjectively assigning projects other projects that are similar in some way.
We can create relationships between projects by calculating something like the classifications/users and looking at nearest neighbours within a range. Obvs need testing to see if this would fluctuate enough. I think if weighted correctly it would.
As relationships change the proximity and connections change and the projects will attract or repel each other.
Q: Would we need to add a continual 'bounce' by oscillation between the last few relationships
Q: What timeframe should the relationship table be updated
Q: how would be motion be determined
See http://getspringy.com/ & http://arborjs.org/

Individual polygons:

Provide feedback