Visuals rearrange #320
Visuals rearrange #320
Conversation
Actual offset/overlap varies by angle, but it exists.
PROBLEM: alternate lines collapse around the NW/SE diagonal
Mulitgraph working at all angular orientations
JosephCottam
added
the
WIP
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
|
This structure was in-part a reaction to difficulties encountered while introducing some network/graph type layouts. (So those are included in this PR as well.) |
JosephCottam
added
awaiting review
|
@marjoleinpnnl I manually merged to the (1) select-exemplar-trajectories, (2) save PNG and (3) click-highlight-legend. Can you add some tests for the the "select-exemplar traces" and "save PNG"? They can add to |
There was a problem hiding this comment.
I have gone through the notebooks and reviewed all the files.
Looks good to me. Thanks @JosephCottam and @marjoleinpnnl for creating such nice visualization tools for us.
* Spring force graph layout and example data. * Marks-nesting of nodes. * Mode reprsented as dounut chart showing attribution sources * Link weight attribution WIP * WIP: First draft rainbow edges Actual offset/overlap varies by angle, but it exists. * Layering and size-relationships improved. * Node labels * Multigraph visual added to the the plots module * Color domain includes links. Color legend added. * Braking apart plots for more clear grouping of functions. * Spring-graph: Label on top of nodes * Mouse manipulation on multi-graph. * Balanced stacking of edges. * Parallel line equation PROBLEM: alternate lines collapse around the NW/SE diagonal * Multigraph working in all orientations (BUT slot is wrong...) Mulitgraph working at all angular orientations * Visual schema facet-check handes post-facet transformation. Added legend to spring-graph * Initial capability to save PNG. * ipy_display produces a png by default. Can do PNG, SVG or interactive. * Highlight legend on chart-click * Select exemplar trajectories by characteristics.
Make the visualization code more maintainable by splitting it into parts.
General principles:
Users should be able to import plots and access most of what they need. Utilities related to specific plot types are segregated into their own files. This makes it easier to identify 'dead code' and do custom data-preprocessing for a broad category of plots without introducing incidental code sharing between plot types.