Conversation
|
After the build completes, the updated documentation will be available here |
Codecov Report✅ All modified and coverable lines are covered by tests.
... and 2 files with indirect coverage changes 🚀 New features to boost your workflow:
|
|
Starts to look great; this will be a very useful addition. |
| ```math | ||
| \sum_g R^{ab}_g F^{abx}_{cg} R^{ax}_c = \sum_{f,h} F^{bax}_{cf} R^{af}_h F^{axb}_{ch} | ||
| ``` | ||
| and the analogous equation with ``a`` and ``b`` swapped. |
There was a problem hiding this comment.
There's more to it than just swapping a and b, but it's probably not worth going into it. Maybe meet halfway and say something about over- and underbraiding?
There was a problem hiding this comment.
yeah, although I guess I really just need to sit down and put some actual figures together 😢
There was a problem hiding this comment.
Any figures that can be borrowed from the TensorKit docs? Or that I can make based on those?
|
Overall amazing work! The pages are pretty short, so I'm wondering if it's beneficial to have this depth to the headers? Also, do you think these docs should also mention the other packages supporting |
|
If we're happy with the current style of the SU2Irrep page, maybe I can help fill in the other pages? It'd be nice to have this available asap, as I've already referred to this PR as documentation twice the past months :) |
Every other representation-category row uses Rep[...]; ZNIrrep was the only one with parentheses. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
'Group Sectors' was ambiguous next to 'Abelian/Non-Abelian Groups' (which are also group-derived, via their irreps). This category holds group-element (Vec_G) sectors, so 'Group Elements' is clearer. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The ZNElement page contrasts itself with ZNIrrep, but the link was one-directional. Add a note distinguishing Rep[Zn] (irreps) from Vec[Zn] (group elements) with a link to ZNElement. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The landing page had only 'Install via the package manager.' and no navigation cues. Add the Pkg add command and links into the Sector Interface and Sector Types sections. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
NamedSector/@NamedSector are exported (src/named.jl) and are a first-class named-tuple counterpart to ProductSector, but had no conceptual page and no table row. Add a page under Composite Sectors mirroring the Product page, wire it into make.jl, and add a table row. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
IsingBimodule is exported but was undocumented anywhere. Add a short 'Multifusion categories' note pointing to it as a reference example, without listing it in the fusion-category summary table. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The block interleaved bare output lines with inputs (not copy-pasteable and not validated). Convert to a jldoctest with julia> prompts so the outputs are checked at build time. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Convert safe LaTeX commands to Unicode inside math contexts (Greek letters, binary operators, relations, arrows, blackboard-bold) for a more readable source. Structural macros (\frac, \sqrt, \overline, matrices, big operators, \ket) and code/doctest blocks are unchanged. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Same Unicode conversion as the previous commit, applied to all sector type pages. Purely a symbol swap inside math contexts; rendering is unchanged (verified by rebuilding the docs) and doctests still pass. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
I did some typo updates and unicode changes, added some missing parts, and think everything is now more or less consistent. Considering the figures, I might actually agree to just have this in a follow-up PR, it seems a bit silly to block all of this for the lack of figures? I'm not sure what you think @borisdevos, but if you want to merge this in its current form, you have green light from me! |
|
Can I take a look tomorrow (= later today). |
I agree, figures can be added at a later time!
Of course, please do! |
Jutho
left a comment
Jutho
left a comment
There was a problem hiding this comment.
I left some small suggestions but otherwise approve.
Co-authored-by: Jutho <Jutho@users.noreply.github.com>
3 participants
This is ongoing work to start properly documenting this package.
It is based on the work in TensorKit, credits go to @Jutho for that.
The docs now consist of two large parts: The first part defines the interface of
Sector, while the second is a list of the concrete types provided by this package.I think the first part fixes #56 and is somewhat ready for review.
I am pretty happy with the general lay-out, but comments are obviously more than welcome.
Many of the formulas will require checking for errors, I have been doing this on the side while waiting for other PRs, so many things will have slipped my full attention.
For the second part, (related to #57) I for now only filled out the
SU2Irreppage in an attempt to show what I had in mind to document the different conventions. Before I continue with the rest (or others can chime in), I wanted to get a round of review in.