Completed in Highcharts v6. Demo: https://www.highcharts.com/demo/parallel-coordinates
Any thoughts on adding the Y-Axis filtering that is done by using a range selection dragged on any of the Y-Axis's? This can be seen in the examples from http://syntagmatic.github.com/parallel-coordinates/
Ta Kernc commented
Here's another Parallel Coordinates implementation with separate yAxes. It's still a hack very much.
Parallel coordinates chart in Highcharts:
This does not have the rich features that the library posted by Kai below has - this is just a proof of concept to show that it is perfectly feasible in Highcharts to do a basic parallel coordinates plot.
Data is the classic 'cars' data set that is often used as the example.
I created a library for parallel coordinates charts here: http://syntagmatic.github.com/parallel-coordinates/
It's similar to line charts, especially since parallel coordinates polylines are rendered the same as line line charts.
But the meaning is encoded differently and interactions are not the same. With line charts you might brush along the X-axis (especially if it's time), or look at a particular data point.
With parallel coordinates, the entire polyline is thought of as a single data point. Interaction is usually brushing any axis (all of which are vertical), reordering axes, or inverting an axis.
Lines show up as points in parallel coordinates (point<->line duality). So in terms of how to interpret it, parallel coordinates is closer to a scatterplot than line chart. When you plot lines in parallel coordinates it looks nothing like a line chart, for instance see this hypercube with a visualization of the edges. The dots in parallel coordinates represent the edges (each edge has a single dot between each axis pair):
The points are the places where polylines intersect, see:
Yes, I understand parallel coordinates :)
They are still just a line chart.
The labeling of the vertical gridlines is not anything out of the ordinary. You basically have a categorized x axis.
At the most extreme, you might need to transform data points to a normalized scale if your different x axis categories each require different y axis scales.
Andreas Müller commented
I think you didn't even understand parallel coordinates...
They have multiple axes - one per category. That´s definetly not the same as in line charts.
parallel coordinates charts are just line charts...