Repository navigation
Regroup the task pages, retire the legacy protocol's RAVEN 2 framing, and more installation/title fixes - #82
Merged
Merged
Conversation
… and more installation/title fixes Third round of feedback on the pydata-sphinx-theme site. - docs/guide/index.md and the 22 task pages: three reconstruction methods (homology, KEGG, context-specific models) were split across two different section headings, with two of them stranded under "Analysis and repair" alongside curation and biomass pages that are really about building and curating a model, not analysing one. Regrouped into Foundations, Simulation, Building and curating, Reconstruction, Analysis and repair, and Comparing and engineering, renumbered 1-22 throughout (headings, subsection numbers, and every cross-reference site-wide), and swapped Deletions and Metabolic tasks within that order. Item 14's list entry dropped "tINIT" from its title; the page's own tINIT-vs-ftINIT comparison is unchanged. - docs/protocol/: "Worked protocol" described the page's role, not its content. Renamed to "GEM reconstruction protocol", numbered subsections (leftover from the source paper's own section numbers) dropped since nothing else on the site uses them, and an info box added stating the original published protocol is RAVEN 2 and this site applies minor modifications for RAVEN 3. Dropped anticipated-results.md, which repeated content already on simulation.md's output. Legacy protocol and legacy tutorials were two separate sidebar sections for what both introductions already call one category; merged into a single "Legacy" section. - docs/installation/raven.md, docs/guide/solvers.md, docs/tutorials/index.md: checkInstallation was renamed to checkRaven in RAVEN 3 (checkInstallation survives only as a deprecated forwarding wrapper); these still called it by the old name. The sample checkRaven output on the installation page was also the old checkInstallation's simpler format, not what checkRaven actually prints (verified against the function's current source). - docs/tutorials/index.md: tutorials 2-4 depend on importExcelModel, which RAVEN 3 removed outright with no forwarding wrapper; nothing on the page said the Excel-editing step no longer runs as written. - docs/installation/data-and-binaries.md: renamed to "Download data and binaries" (its own contents are download instructions, not a record of what has already been downloaded). "Where it goes" gave one cache path and called it universal, when raven-toolbox's cache directory resolves to a different absolute path on each OS; added the three real paths, verified against raven-toolbox's own cache-path source, each on its own tab. Dropped "Integrity and versions" and "Licences", which repeated general package-management background unrelated to this site's own setup steps. - docs/raven3-vs-raven-toolbox.md: the solvers table said MATLAB's GLPK comes via the COBRA Toolbox; it ships bundled with RAVEN itself, no COBRA Toolbox involved. "Moving a model between them" and "Coming from RAVEN 2.0 to Python" repeated the YAML/SBML interoperability and RAVEN-2-specific background covered in more depth on yaml-format.md and raven3-migration.md respectively. Retitled to "RAVEN and raven-toolbox" (RAVEN 3 is the only supported line now) and added an early pointer to the MATLAB vs Python table for readers looking for one specific function. - docs/raven3-migration.md: retitled to "RAVEN 2 to RAVEN 3" for the same reason. - docs/matlab-vs-python.md, docs/api/index.md, docs/migrate.md: the mapping table is a function reference, not a migration path; moved it from under Migrate to under API reference. - docs/references.md: the INIT and ftINIT citations sat under one heading with a note explaining they were different papers; split into their own "ftINIT" and "tINIT" subsections so each is under its own citation. - docs/index.md: the MATLAB (Add-Ons) install step ran its two instructions into one paragraph; split onto separate lines. The Python (git) install cloned the full history where the MATLAB (git) tab already used --depth=1.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Third round of feedback on the pydata-sphinx-theme site.
(homology, KEGG, context-specific models) were split across two different
section headings, with two of them stranded under "Analysis and repair"
alongside curation and biomass pages that are really about building and
curating a model, not analysing one. Regrouped into Foundations,
Simulation, Building and curating, Reconstruction, Analysis and repair,
and Comparing and engineering, renumbered 1-22 throughout (headings,
subsection numbers, and every cross-reference site-wide), and swapped
Deletions and Metabolic tasks within that order. Item 14's list entry
dropped "tINIT" from its title; the page's own tINIT-vs-ftINIT comparison
is unchanged.
content. Renamed to "GEM reconstruction protocol", numbered subsections
(leftover from the source paper's own section numbers) dropped since nothing
else on the site uses them, and an info box added stating the original
published protocol is RAVEN 2 and this site applies minor modifications for
RAVEN 3. Dropped anticipated-results.md, which repeated content already on
simulation.md's output. Legacy protocol and legacy tutorials were two
separate sidebar sections for what both introductions already call one
category; merged into a single "Legacy" section.
checkInstallation was renamed to checkRaven in RAVEN 3 (checkInstallation
survives only as a deprecated forwarding wrapper); these still called it by
the old name. The sample checkRaven output on the installation page was
also the old checkInstallation's simpler format, not what checkRaven
actually prints (verified against the function's current source).
RAVEN 3 removed outright with no forwarding wrapper; nothing on the page
said the Excel-editing step no longer runs as written.
binaries" (its own contents are download instructions, not a record of
what has already been downloaded). "Where it goes" gave one cache path
and called it universal, when raven-toolbox's cache directory resolves to
a different absolute path on each OS; added the three real paths,
verified against raven-toolbox's own cache-path source, each on its own
tab. Dropped "Integrity and versions" and "Licences", which repeated
general package-management background unrelated to this site's own setup
steps.
comes via the COBRA Toolbox; it ships bundled with RAVEN itself, no COBRA
Toolbox involved. "Moving a model between them" and "Coming from RAVEN 2.0
to Python" repeated the YAML/SBML interoperability and RAVEN-2-specific
background covered in more depth on yaml-format.md and
raven3-migration.md respectively. Retitled to "RAVEN and raven-toolbox"
(RAVEN 3 is the only supported line now) and added an early pointer to the
MATLAB vs Python table for readers looking for one specific function.
reason.
table is a function reference, not a migration path; moved it from under
Migrate to under API reference.
with a note explaining they were different papers; split into their own
"ftINIT" and "tINIT" subsections so each is under its own citation.
into one paragraph; split onto separate lines. The Python (git) install
cloned the full history where the MATLAB (git) tab already used
--depth=1.