Updated docs with snntorch to Norse example and dev docs (#109)

* Updated docs with snntorch to Norse example and dev docs

* Updated docs for porting

---------

Co-authored-by: jegp <jens@jepedersen.dk>

docs/source/_config.yml

@@ -4,4 +4,13 @@ logo: ../logo_light.png
 
 repository:
   url: https://github.com/neuromorphs/nir
-  branch: main
+  branch: main
+  path_to_book: docs/source
+
+execute:
+  execute_notebooks: off
+
+launch_buttons:
+  notebook_interface: "jupyterlab"
+  binderhub_url: "https://mybinder.org/v2/gh/neuromorphs/nir/main?urlpath=lab"
+  colab_url: "https://colab.research.google.com"

docs/source/_toc.yml

@@ -24,8 +24,9 @@ parts:
   - file: examples/snntorch/nir-conversion
   - file: examples/spinnaker2/import
   - file: examples/spyx/conversion
+  - file: examples/snntorch_to_norse
 - caption: Developer guide
   chapters:
+  - file: porting_nir
   - file: api_design
-  - file: dev
-    title: Contributing
+  - file: contributing

docs/source/api_design.md

@@ -1,6 +1,6 @@
 # API design
 
-The reference implementation simply consists of a series of Python classes that *represent* [NIR structures](primitives).
+NIR is simple: it consists of a series of objects that *represent* [NIR structures](primitives).
 In other words, they do not implement the functionality of the nodes, but simply represent the necessary parameters required to *eventually* evaluate the node.
 
 We chose Python because the language is straight-forward, known by most, and has excellent [dataclasses](https://docs.python.org/3/library/dataclasses.html) exactly for our purpose.
@@ -16,6 +16,14 @@ In this example, we create a class that inherits from the parent [`NIRNode`](htt
 Instantiating the class is simply `MyNIRNode(np.array([...]))`.
 
 ## NIR Graphs and edges
+```{figure} nir_graph_example.svg
+---
+height: 200px
+name: nir-graph-example
+---
+An example of a NIR graph with four nodes: Input, Leaky-Integrator, Affine map, and Output.
+```
+
 A collection of nodes is a `NIRGraph`, which is, you guessed it, a `NIRNode`.
 But the graph node is special in that it contains a number of named nodes (`.nodes`) and connections between them (`.edges`).
 The nodes are named because we need to uniquely distinguish them from each other, so `.nodes` is actually a dictionary (`Dict[str, NIRNode]`).