Document Transformations
#73
Conversation
Codecov Report
@@ Coverage Diff @@
## master #73 +/- ##
==========================================
- Coverage 78.13% 77.82% -0.32%
==========================================
Files 10 9 -1
Lines 709 699 -10
==========================================
- Hits 554 544 -10
Misses 155 155
|
mofeing
changed the title
transformations.jlTransformations
There was a problem hiding this comment.
Great work! Just a couple of comments:
- It's best practice to use the infinitive for explaining functions in the documentation (e.g. use "Reduce" instead of "Reduces". Imagine that function are like verbs or actions).
- One of the main reasons why these methods were developed is because it drastically reduces the problem size of contraction path search. We should mention it.
- Reducing the maximum rank of the Tensor Network is not what it solves, but reducing the rank (or better, size) of the involved tensors.
- It would a good idea to cite the paper by Johnnie Gray and maybe say that we have taken inspiration? from
quimb. - I think it would be great to put small examples inside the docstrings for each
Transformation, so users can really appreciate each of them.
docs/src/transformations.md
Outdated
| # Examples | ||
| ```julia-repl | ||
| julia> tn = TensorNetwork(...) | ||
| julia> transformed_tn = transform(tn, Tenet.RankSimplification) | ||
|
|
||
| julia> fig = Figure() | ||
| julia> ax1 = Axis(fig[1, 1]; title="Original TensorNetwork") | ||
| julia> p1 = plot!(ax1, tn; node_size=5.) | ||
| julia> ax2 = Axis(fig[1, 2], title="Transformed TensorNetwork") | ||
| julia> p2 = plot!(ax2, tn2; node_size=5.) | ||
| julia> ax1.titlesize=20 | ||
| julia> ax2.titlesize=20 | ||
| ``` | ||
| ```@raw html | ||
| <figure> | ||
| <img width=500 src="../assets/transformation.svg" alt="Before and after transformation in a Tensor Network"> | ||
| </figure> | ||
| ``` | ||
|
|
There was a problem hiding this comment.
I prefer to put examples inside the docstrings of the different Transformations, so the user can check their effect.
(Use small examples)
There was a problem hiding this comment.
Should we show a visual representation for each transformation? Maybe this would be too much
There was a problem hiding this comment.
I think this would be a good idea, but using very small tensor networks. Similar to what they do in the paper with the diagrams.
There was a problem hiding this comment.
I can help with this if you need help.
There was a problem hiding this comment.
I will try to create some examples, I will let you know!
Co-authored-by: Sergio Sánchez Ramírez <[email protected]>
Co-authored-by: Sergio Sánchez Ramírez <[email protected]>
Co-authored-by: Sergio Sánchez Ramírez <[email protected]>
|
@mofeing thanks for the comments, now the docs work as expected |
|
I prefer multiple smaller examples and leave the RQC for the examples, but it's good enough right now. I will change it in the future. |
Summary
This PR aims to enhance the documentation for the
transformations.jlmodule, providing more detailed descriptions of each available transformation and including a visual example to demonstrate the impact of the transformations on a tensor network.