F#

Based on pythonnet/pythonnet#554

> #r C:\\Python\\Python35_64b\\lib\\site-packages\\Python.Runtime.DLL
> using Python.Runtime;
> dynamic np; using (Py.GIL()) { np = Py.Import("numpy");};
> using System.Collections.Generic;
> using System.Collections;
> dynamic a = np.array(new List<float> { 1, 2, 3 });
> dynamic b = np.array(new List<float> { 6, 5, 4 }, dtype:

This is to track implementation of the ML-Features: https://spark.apache.org/docs/latest/ml-features

Bucketizer has been implemented in dotnet/spark#378 but there are more features that should be implemented.

• Feature Extractors
  • TF-IDF
  • Word2Vec (dotnet/spark#491)
  • CountVectorizer (https://github.com/dotnet/spark/p

In Statistics module, we do calculations and it would be beneficial to cover it by unit or property-based tests.

Here you can take a look at examples with property-based testing.

Came up in #526

Adding this would bring the NUnit integration to feature parity.

This was already proposed in #93, but given how much the landscape has changed since 2013, I thought it better to create a new issue.

The specific reason why I'm proposing it right now is to be able to use FSharp.Data in WebAssembly with Bolero; see this issue. But more generally I think it would be good to update to the modern and recommended li