Context tutorial (#481)

* Add context tutorial

* Add more detail about contexts to compiler design page

* Apply suggestions from code review

Co-authored-by: Hong Ge <[email protected]>

* Use explicit imports in contexts

* Remove unnecessary deps for contexts

* Update page titles to match ToC

* Improve a paragraph in context tutorial

* Small fixes to contexts tutorial

* Apply review feedback

Co-authored-by: Will Tebbutt <[email protected]>

* Apply review feedback

Co-authored-by: Will Tebbutt <[email protected]>

* Clarify a sentence in context tutorial

---------

Co-authored-by: Hong Ge <[email protected]>
Co-authored-by: Will Tebbutt <[email protected]>

Author: 3 people

_quarto.yml

@@ -96,8 +96,10 @@ website:
               - text: "How to Contribute"
                 href: tutorials/docs-01-contributing-guide/index.qmd
               - tutorials/docs-05-for-developers-compiler/index.qmd
-              - text: "A Mini Turing Compiler"
+              - text: "A Mini Turing Implementation I: Compiler"
                 href: tutorials/14-minituring/index.qmd
+              - text: "A Mini Turing Implementation II: Contexts"
+                href: tutorials/16-contexts/index.qmd
               - tutorials/docs-06-for-developers-interface/index.qmd
 
 

tutorials/14-minituring/index.qmd

@@ -1,5 +1,5 @@
 ---
-title: MiniTuring
+title: "A Mini Turing Implementation I: Compiler"
 engine: julia
 ---
 
@@ -81,8 +81,8 @@ For instance, in some cases we might want to sample all variables from their pri
 Thus depending on the inference algorithm we want to use different `assume` and `observe` implementations.
 We can achieve this by providing this `context` information as a function argument to `assume` and `observe`.
 
-**Note:** *Although the context system in this tutorial is inspired by DynamicPPL, Turing's context system is much more complicated for flexibility and efficiency reasons.
-Thus readers are advised to refer to the documentation of DynamicPPL and Turing for more detailed information about their context system.*
+**Note:** *Although the context system in this tutorial is inspired by DynamicPPL, it is very simplistic.
+We expand this mini Turing example in the [contexts](../16-contexts) tutorial with some more complexity, to illustrate how and why contexts are central to Turing's design. For the full details one still needs to go to the actual source of DynamicPPL though.*
 
 Here we can see the implementation of a sampler that draws values of unobserved variables from the prior and computes the log-probability for every variable.
 
@@ -290,4 +290,4 @@ end
 sample(turing_m(3.0), MH(ScalMat(2, 1.0)), 1_000_000, progress=false)
 ```
 
-As you can see, with our simple probabilistic programming language and custom samplers we get similar results as Turing.
+As you can see, with our simple probabilistic programming language and custom samplers we get similar results as Turing.