Crime-sim-toolkit

Introducing the Microsimulator class

This release now includes the `crime_sim_toolkit.microsim.Microsimulator` class that allows for simulating a crime reports for individuals in a synthetic population from transition probabilities calculated from historic crime data.

Also released via [Python package index](https://pypi.org/project/crime-sim-toolkit/)

Changelog

* Added microsim.Microsimulator class
* Separated tests into separate test files for each module
* Added github actions for this repo
* Added new util function that reverse maps Crime_descriptions from populate_offence into 15 [PoliceUK data](https://data.police.uk/) classes
* Added new example notebooks for the Microsimulator class in examples/
* Updated the README to account for new changes and more explicit project description
* Within examples added HPC/sge_scripts outlining use examples for HPC
* Created [GitHub wiki](https://github.com/Sparrow0hawk/crime_sim_toolkit/wiki) for this project
* **BUG FIX** refactored `utils.populate_offence` function to remove column name adjustment, thus making the function more flexible for inputs not generated by the `Poisson_sim`

It was discovered that using utils.sample_perturb returns counts as floats not ints leading to problems downstream in utils.counts_to_reports because floats are not iterable.

This was resolved in pull request 36

Includes datetime compatibility, a moving window for sampling, and an additional util sample_perturb that can up or down sample models.

Running a release of current work in a useable capacity via PyPi and GitHub releases.