When naming this gem, I had in mind the German level a.k.a spirit level. But naming was not quite in the beginning of the development process. All in the beginning was the wish to create an algorithm to automatically draw trend lines. So for like 2 year I was whirling my head on how to put an algorithmic ruler on a chart and rotate it until one or a set of trend lines are concise outpout.

As most hard to accomplish things turn out to be much easier when you put them up side down, same happened to me in this matter. I found rotating the ruler is too hard for, but instead transforming (shearing) the chart itself while keeping the ruler at its level is much more eligible.

The idea and the development of the algorithm had taken place within another Cotcube project named 'SwapSeeker'. At some point the SwapSeeker has become too complex so I decided to decouple the Level and the Stencils as separate functional unit, that can be used on arbitrary time series.

There is really no magic in it. The timeseries (or basically an interval of a timeseries) needs to be prepared to locate in Cartesian Quadrant I with fitting x==0 to y==0, and then a binary search on shearing angles determines the resulting muiltitangent of which the origin ('now') as one point already is given. One limitation applies: Only shearing between 0 and 90 degrees is supported.

The result contains

• only the origin, if it is the absolute high (resp. low) within the provided interval of the time series. This happens to happen if deg -> 0.
• the origin and one arbitrary point, the usual result.
• the origin and two or more points residing on the same level--what is the desired result showing mathematical accuracy where human eye cannot detect it in time.

The 'tri' is about that to find an artificially supported slope (i.e. a swap) at least 3 points have to reside on it. It is leaned to triangulate, but instead of working with angles we are working with tangents here.

Shear mapping and slope detection might find a slope meeting the requirements. But in tritangulate, near misses are considered and added as members.

The shearing transformation is based on x and y values, where y obviously are the values of the series while x refers to the time. It turned out a much bigger challenge to create an expedient mapping from DateTime to Integer, where, as you foresee, 'now' should result in x.zero?.

TODO: Write usage instructions here

After checking out the repo, run bin/setup to install dependencies. Then, run rake spec to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.

To install this gem onto your local machine, run bundle exec rake install. To release a new version, update the version number in version.rb, and then run bundle exec rake release, which will create a git tag for the version, push git commits and tags, and push the .gem file to rubygems.org.

The gem is available as open source under the terms of the MIT License.