Quiver timeline, diachronic view - color coding dependent on metric

[MareenGeestmann]

Assuming color coding as

• green: better

• yellow: equal

• red: worse

in comparison to last release, then it differs between metrics. E. g. for "Pages per minutes" higher values are better while for "CER Mean" lower values are better.

Hence, color coding must be adapted dependent on the metric.

[MareenGeestmann]

The color changes when enlarging the plot. E.g. Pages per Minute, first average plot.
How so?

[paulpestov]

Fixed

[MareenGeestmann]

Fine!

New discovery: In CER standard deviation for two documentes the average is green (values lower) or red (values higher) but all workflows are yellow (values equal). How so?
(Bettina von Arnim. Tagebuch. 1835.;
Alberti, Leon Battista. De pictura...)

=> In fact, the source lies in the data, not in a false coloring. So I open a new issue for it.