syntagmatic/parallel-coordinates

D3 V4 Examples

syntagmatic opened this issue · 13 comments

This project is too complex with too many outstanding issues for me to port to V4, but I have implemented some parallel coordinates examples with several of the features from this library and some enhancements to the core pattern described in #91

Here are the two examples:

Exoplanets II
https://bl.ocks.org/syntagmatic/482706e0638c67836d94b20f0cb37122

Nutrient Parallel Coordinates IV
https://bl.ocks.org/syntagmatic/05a5b0897a48890133beb59c815bd953

Also check out this repo for a port by @mlarosa2 if you need features from the library in v4 https://github.com/mlarosa2/parcoords

See #326 for past discussion

I've ported parcoords to D3 V4 and ES6 to a project called parcoords-es .
All original parcoords examples have been ported and verified.

It is totally OK to use parcoords with d3 v4 modules. However, I create parcoords-es because I do need a pure D3 V4 version of this visualization in my work.

Thank you for creating this visualization!

@BigFatDog wow man. This is awesome. Just a quick question. Did you try using this with TS?

No, I didn't. I haven't seriously considered TS in my visualizations yet.

Here is a simpler version of ordinal and numerical parallel coordinate in d3 V4.
Is there a way to brush ordinal axis as well ; like the "name " column in this example
http://plnkr.co/edit/TiM6ZsvMTTBhh8ZdMoFQ?p=preview

Hello ,

Can anyone please share a little simpler parallel coordinate example in d3 V4 , that has both ordinal and numerical axis. Also basic features like brushing , brush snapping and dragging .

Any help would be highly appreciated.
Thanks

@syntagmatic,

Is it possible to direct users to @BigFatDog or other projects that work with the current D3 (v4)? Perhaps a link in the main readme?

This project is great and works well - I just had many struggles until I found this issue and saw that people have already done a port.

@syntagmatic
I'm actively maintaining a d3v4 port of this project. Would you please consider adding my project parcoords-es to a D3V4 list in your readme?

Yun Xing

Hello,
I was referring to exoplanets example of parallel coordinate using d3 V4 https://bl.ocks.org/syntagmatic/482706e0638c67836d94b20f0cb37122

The issue I am observing is that when we brush the log scales (like Planet Semi major-axis or Planet orbital period) part of the graph appears to be broken in the middle.

Here is the screen shot of this .
https://user-images.githubusercontent.com/32777204/40357473-8ef90686-5dbc-11e8-8d64-cbaf3349cac0.png

Could anyone please tell me what is the correct interpretation of that !
Thanks

@Abhijeet-Das In that example missing/null values cause broken lines. The horizontal lines in the Mass in Jupiters axis, for instance, indicate that data point has no values for Planet Inclination or Planet Radius.

In previous examples, the polyline was just directly connected to axes that did have data points, passing through dimensions with missing data. It's a misleading way to render missing data, since straight lines across multiple axes are very significant in parallel coordinates.

@syntagmatic , Thanks a lot for your quick response and clarification. I guess I should add some default values in my graph for the missing ones to make it appear consistent. Thanks again

I added a link to @BigFatDog's https://github.com/BigFatDog/parcoords-es in the readme and noted the library is no longer in active development in the Github project description.