Skip to content

Releases: curran/d3-area-label

v1.6.0

01 Mar 14:31
Compare
Choose a tag to compare

What's Changed

Full Changelog: v1.5.0...v1.6.0

All methods chainable

16 Jul 08:00
Compare
Choose a tag to compare

Thanks to @Legion4444 for this contribution that makes all methods chainable! Details in #24

Also upgraded dependencies to the latest versions.

Enjoy!

ES6 Fix & Change in interpolateResolution Default

16 Aug 10:47
Compare
Choose a tag to compare
  • ES6 Fix makes this library work well with Webpack (#19)
  • Changed default interpolateResolution from 200 to 800 based on observation that 200 is not enough usually.

Interpolation

03 Aug 19:28
Compare
Choose a tag to compare

Use Interpolation #16

Before (in Refugees StreamGraph):
image

After:
image

Before (in test/smallN.html):
image

After:
image

Correctness

03 Aug 06:43
Compare
Choose a tag to compare

Guarantee Labels are Inside Areas #14.

There was an issue where when the gaps between data points was large, the labels were not always ending up inside the area.

Before:
image

After:
image

1.0

29 Jul 14:31
Compare
Choose a tag to compare
1.0
  • API change - function now returns an object with a toString method that generates an SVG transform.
  • Accessors #6
  • Padding #8
  • Algorithm improvements:
    • Compute maxHeight #9
    • Break inner test loop as soon as we know the rectangle will not fit
  • Expose number of iterations and raw coordinates in result

Algorithm improvement

29 Jul 06:27
Compare
Choose a tag to compare

Break out of the inner test loop as soon as we know the rectangle will not fit.

Bisection Method

27 Jul 12:08
Compare
Choose a tag to compare

This version uses the Bisection Method to search for the largest size label, using fewer iterations than before, and achieving more accurate results.

Now the algorithm uses a tolerance for accuracy of the height of the label bounding box in pixels, which is by default 0.01. Using this tolerance, the bisection method takes around 18 iterations on average to find a solution, as compared to on average around 100 with the previous linear search method, which was only accurate to a tolerance of 1.

First Working Version

27 Jul 10:38
Compare
Choose a tag to compare

image

The algorithm is not efficient (linear scan of possible sizes), and the results are not precise (integer height increments are searched only), but it's basically working.