Milesgranger Gap Statistic Versions Save

• User can pass in their own clustering function into OptimalK (https://github.com/milesgranger/gap_statistic/commit/fe1b1d31c93c592eb597474493f520b7f6b6434c)
• Switch to Azure Pipelines (https://github.com/milesgranger/gap_statistic/commit/9020d9530d15e1ed79045e25e1ca5e6c67fa5c0c, https://github.com/milesgranger/gap_statistic/commit/db030eea4e362acc4a4d43029c7ddb701ff24c95)
• Use black formatting for the project (https://github.com/milesgranger/gap_statistic/commit/bdb4af1467533f3d8091d05c83012ea040f0ffa9)

Release contains an initial stable release of the same API found in the first release, along with OptimalK taking an additional value "rust" for the backend parameter.

Initial release including a Rust backend implementation of the gap statistic & kmeans.

Support for Joblib, Multi-processing and single core