two-coins.Rmd

---
title: "Inference on two coins with Infer.NET"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{matchmaking-infer-net}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
```

This vignette demonstrates the `dotnet` knitr engine for use in R Markdown chunks that have C# and F# code. First, register the engine with knitr:

```{r setup}
dotnet::register_engine()
```

The example used comes from the [two coins tutorial](https://dotnet.github.io/infer/userguide/Two%20coins%20tutorial.html). In this example, which uses the "Microsoft.ML.Probabilistic.Compiler" NuGet package, `dotnet add package Microsoft.ML.Probabilistic.Compiler` can be accomplished by specifying it in `add_packages` inside `engine.opts` chunk option:

````{verbatim}
```{dotnet, engine.opts = list(add_packages = c('Microsoft.ML.Probabilistic.Compiler'))}
using System;
using Microsoft.ML.Probabilistic.Models;
```
````

And here is the probabilistic program written in C#:

```{dotnet csharp, engine.opts = list(add_packages = c('Microsoft.ML.Probabilistic.Compiler')), cache = TRUE}
using System;
using Microsoft.ML.Probabilistic.Models;

class Program
{
  static void Main(string[] args)
  {
    Variable<bool> firstCoin = Variable.Bernoulli(0.5);
    Variable<bool> secondCoin = Variable.Bernoulli(0.5);
    Variable<bool> bothHeads = firstCoin & secondCoin;
    
    // Inferring distributions:
    InferenceEngine engine = new InferenceEngine();
    Console.WriteLine("Probability both coins are heads: " + engine.Infer(bothHeads));
    
    // Backwards reasoning:
    bothHeads.ObservedValue = false;
    Console.WriteLine("Probability distribution over firstCoin: " + engine.Infer(firstCoin));
  }
}
```