docs: finish hubbard Brinkmann-Rice tutorial

.github/actions/setup-triqs/action.yml

@@ -8,10 +8,10 @@ inputs:
     default: "3.10"
     required: false
   cc:
-    default: "clang"
+    default: "gcc"
     required: false
   cxx:
-    default: "clang++"
+    default: "g++"
     required: false
   llvm:
     default: "15"
@@ -120,11 +120,12 @@ runs:
       cmake .. -DCMAKE_INSTALL_PREFIX=$HOME/triqs_install -DBuild_Documentation=OFF -DBuild_Tests=OFF
       make -j1 VERBOSE=1
       cd ../../
-      echo "TRIQS_INSTALL=$HOME/triqs_install" >> $GITHUB_ENV
 
   - name: Install TRIQS
     shell: bash
-    run: cd triqs/build && make install && cd ../../
+    run: |
+      cd triqs/build && make install && cd ../../
+      echo "TRIQS_INSTALL=$HOME/triqs_install" >> $GITHUB_ENV
 
 branding:
   icon: "package"

.github/workflows/docs.yml

@@ -2,7 +2,6 @@ name: Docs
 
 on:
   workflow_dispatch:
-  pull_request:
   push:
     branches:
       - main

TODO.md

@@ -18,14 +18,19 @@ version 3.2.0
 - Helper plot functions
 - Helper functions for calculating free/total energy
 - add linting actions
+- pipy packaging
+- dimer insulator to explain some different insulating states
+- change references in docs to use sphinxcontrib-bibtex
+
 
 ### In Progress
 
 - symmetry representation tutorial
 - Kagome/projectors tutorial
-- single-orbital Hubbard tutorial 
 - Finish TBmodels in tight-binding how-to
 - Sort out cross-referencing to API in docs
+- action for hatchling version bump
 
 ### Done ✓
-- Setup github actions
+- Setup github actions
+- single-orbital Hubbard tutorial 

docs/how-to/embedding.md

@@ -8,7 +8,7 @@ bit about how they do things.
 [About the embedding Hamiltonian](../explanations/embedding.md).
 :::
 
-## :py:class:`EmbeddingAtomDiag`
+## `EmbeddingAtomDiag`
 
 First a `gf_struct` object has to be created that describes the correlated 
 subspace $\mathcal{C}_i$. This object should be constructed in the same way 
@@ -81,6 +81,17 @@ where `h0_loc_matrix` is a matrix that describes the couplings, with the same
 block matrix structure as `Lambda_c` and `D`. This is stored in 
 `embedding.h0_loc` as a {{TRIQS}} operator.
 
+### Setting `h_int`
+
+If you want to update the interaction terms on the impurity 
+
+```python
+# A new h_int
+h_int = ...
+
+embedding.set_h_int(h_int)
+```
+
 ### Solving
 
 It is solved as
@@ -144,7 +155,7 @@ embedding.gf_struct_emb
 ```
 
 
-## :py:class:`EmbeddingDummy`
+## `EmbeddingDummy`
 
 If you want to have some correlated subspaces $\mathcal{C}_i$ as inequivalent,
 but they are related by some symmetry, it is not necessary to solve for the