Skip to content

Round rank output, modernise the docs, add diagrams - #121

Merged
neksa merged 3 commits into
mainfrom
docs/theme-and-precision
Sep 8, 2026
Merged

neksa merged 3 commits into
mainfrom
docs/theme-and-precision

Conversation

@neksa

@neksa neksa commented Sep 8, 2026

Copy link
Copy Markdown
Owner

rank printed 8.09229314668869e-08 — fifteen significant figures for an estimate from a binomial test on single-digit counts. Now four:

gene     mutation  mutability  observed  bscore     qvalue     label
BOD1L    T2810S    8.092e-08   1         3.561e-06  5.453e-05  Driver

Two bugs found while checking that:

  • read_protein_mutations_MAF aborted the whole file on one row with the wrong field count, so rank died on tests/motifs/data/tcga_A3JM-01.maf — a file motif and profile read fine. The other readers learned to skip such rows in 1.1.0; this one hadn't. With it fixed, that fixture ranks CTNNB1 S37C and TP53 R248P.
  • A ranking producing no rows wrote an empty file in silence. It now says the input needs protein annotations.

Docs: Furo theme, and mermaid diagrams where prose was doing badly — how the subcommands chain, the FILTER decision, the cohort argument precedence, and the re-run swap. The rank examples showed the old precision, so they now show real output.

Labelled minor: the rank output format changes.

The numeric columns were written at full float precision, so a mutability of
8.09229314668869e-08 claimed a precision that a binomial test on single-digit
counts does not have, and made the table hard to read. Four significant figures.

read_protein_mutations_MAF also aborted the whole file on one row with the wrong
number of fields, so rank failed on real TCGA files that motif and profile read
without complaint; the other readers learned to skip such rows in 1.1.0. And a
ranking that produced no rows wrote an empty file in silence rather than saying
the input needs protein annotations.

The examples in rank_doc showed the old precision, so they now show the real
output.
Switches to Furo and adds mermaid diagrams where prose was doing badly: how the
subcommands chain together, the FILTER decision, the precedence of the
overlapping cohort arguments, and the re-run swap.

The rank examples showed fifteen-digit numbers the command no longer prints, so
they now show its actual output, and the tables they sit in are rebuilt to
match.
@neksa neksa added the minor Minor version bump (0.X.0) label Sep 8, 2026
The check grepped for 'MutaGene X.Y.Z documentation', which is alabaster's
title. Furo uses html_title, so changing the theme read as a version mismatch.
It matches the version alone now.
@neksa
neksa merged commit 88d6463 into main Sep 8, 2026
11 checks passed
@neksa
neksa deleted the docs/theme-and-precision branch September 8, 2026 21:48
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

minor Minor version bump (0.X.0)

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant