The implemented `FisherJenks` classifier is very slow. I would like to suggest using [jenkspy](https://github.com/mthh/jenkspy) instead. It's written in `cython` and it's very fast.
