Tutorial

CHANGELOG.md

@@ -5,6 +5,7 @@
 ## Current develop
 
 ### Added (new features/APIs/variables/...)
+- [[PR 1439]](https://github.com/parthenon-hpc-lab/parthenon/pull/1439) Long-overdue parthenon tutorial
 - [[PR 1443]](https://github.com/parthenon-hpc-lab/parthenon/pull/1443) Allow for specifying topological elements in var views
 - [[PR 1431]](https://github.com/parthenon-hpc-lab/parthenon/pull/1431) Add reductions to the loop abstraction
 - [[PR 1429]](https://github.com/parthenon-hpc-lab/parthenon/pull/1429) Fixes to swarm tensor xdmf

doc/sphinx/src/README.rst

@@ -4,6 +4,12 @@ Overview
 This file provides an overview of different (not necessarily all)
 features in Parthenon and how to use them.
 
+Tutorial
+---------
+
+See our :ref:`tutorial <tutorial>` for a gentle introduction to building your own
+Parthenon-based AMR code.
+
 Building parthenon
 ------------------
 

doc/sphinx/src/chapters/getting_started.rst

@@ -9,6 +9,7 @@ so you can run your first example quickly.
    :maxdepth: 1
 
    ../README
+   ../tutorial
    ../building
    ../inputs
    ../outputs

doc/sphinx/src/loop_abstraction.rst

@@ -300,7 +300,7 @@ values a consumer loop needs. If a consumer inner loop runs over a logical point
 ``S``, then a producer that fills scratch for the consumer must cover ``S`` plus the
 shifted copies named by the halo:
 
-.. code:: cpp
+.. code::
 
    AddHalo<halo_t<h1, h2, ...>>(S) == S ∪ shift(S, h1) ∪ shift(S, h2) ∪ ...
 
@@ -325,9 +325,9 @@ the same index conventions as the loop body:
   requested on construction.
 
 Each view accepts the same index forms the body produces (flat ``int``, ``Index3``,
-or explicit ``k, j, i``), so a kernel can be written once and reused across inner
-tags. In `inner_tag::logical_coords` loops, these are just light wrappers that call
-through to the sparse packs. For all other `inner_tag`s, pack view construction directly
+or explicit ``k``, ``j``, ``i``), so a kernel can be written once and reused across inner
+tags. In ``inner_tag::logical_coords`` loops, these are just light wrappers that call
+through to the sparse packs. For all other ``inner_tag`` s, pack view construction directly
 pulls out pointers to the variables. This can promote vectorization and be a significant
 performance benefit. 
 

doc/sphinx/src/par_for.rst

@@ -1,4 +1,4 @@
-.. par_for:
+.. _par_for:
 
 Parallelism
 ===========