pirl-unc / tcrsift / 25583981870 / 3
71%
main: 71%

DEFAULT BRANCH: main
Ran
Files 24
Run time 1s
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 70.972% (+0.2%) from 70.81%
25583981870.3

push

github

web-flow 
Add per-axis clonotype aggregations for timepoint/APC/tissue (#9 ch2) (#19)

Second slice of #9. Same shape as #8 chunk 2 but generalized to the
new axes registered in chunk 1 (v0.6.0).

In aggregate_clonotypes, after the existing donor/method block:

  - timepoints / n_timepoints      — when timepoint axis populated
  - apcs / n_apcs                  — when apc_type axis populated
  - tissues / n_tissues            — when tissue axis populated
  - timepoints_per_donor (JSON)    — when both donor and timepoint set
    + max_timepoints_per_donor       backs --min-timepoints-per-donor
  - apcs_per_donor (JSON)          — when both donor and apc_type set
    + max_apcs_per_donor             backs --min-apcs-per-donor

JSON encoding for the dict-shape columns matches the methods_per_donor
convention from #8 ch2 — clean h5ad/CSV serialization, easy round-trip
with json.loads / pd.read_json.

All blocks emit conditionally on the relevant sample-sheet fields
being supplied. Designs that don't use the new axes get no extra
columns — preserves the optional-metadata pattern from #5.

Per-axis cell-count and frequency aggregations (analogous to
max_cells_per_method / max_frequency_per_method from #15 ch3) are
not included here. They'd only be needed if chunk 3's filter knobs
ask for --min-cells-per-timepoint etc.; deferring keeps this PR
focused on the issue's chunk-2 ask.

Tests cover:
  - all single-axis lists / counts populated correctly
  - JSON dict shape and max_*_per_donor values for the canonical
    "shared across timepoints / APCs within donor" case
  - per-donor nested aggregations only emitted when both axes set
  - all axis columns absent when no axis fields are set

Bumps version to 0.7.0 — new columns on the clonotype-table schema
(additive, no breaking change).

Next chunk: filter knobs (--min-timepoints-per-donor,
--min-apcs-per-donor, etc.) that read these columns.

4220 of 5946 relevant lines covered (70.97%)

0.71 hits per line

Source Files on job python-3.10 - 25583981870.3