Skip to content

docs: more pedagogic tutorials for data_organization and condition_comparisons - #185

Open
saatvik-viniak wants to merge 27 commits into
NygenAnalytics:masterfrom
saatvik-viniak:stats_testing
Open

docs: more pedagogic tutorials for data_organization and condition_comparisons#185
saatvik-viniak wants to merge 27 commits into
NygenAnalytics:masterfrom
saatvik-viniak:stats_testing

Conversation

@saatvik-viniak

@saatvik-viniak saatvik-viniak commented Sep 5, 2026

Copy link
Copy Markdown
Contributor

Makes two tutorials more intuitive and pedagogic (draft right now, slowly working on things).

data_organization was the first section

  • Refresher with Scanpy/Seurat equivalents
  • Checkpoints per section plus revert boxes: live-I backup/restore with update_key, reset_key, insert/drop demos (executed) to teach the user how to go back if needed
  • Immutable-selection notes for filter_cells, revert cheat-sheet, grouped mistakes

condition_comparisons was the second thing I focused on

  • Crash course on cell-specific stats testing, adapted to the donor-level paired design: independence/pseudoreplication limits, sparsity and exchangeability, rare-population and multimodality benefits with caveats. This section requires the most background, I'd say, so you know which stats to run and the pros and cons. By prefacing with this right away, the user builds their foundation.

saatvik-viniak and others added 27 commits August 6, 2026 22:05
…re functional now, more things to come soon.
…'stacked_violin' and scope='panel', the color bar will now show relative expression of a gene inside of its individual panel inside of the entire plot, thus it makes things easier in a sense for comparison qualitatively what gene is the highest expressed inside of a cluster.
Hardening changes and doing some revisions for the non-parametric tests.
… knobs; reuse validator checks persisted fingerprints; group mask preserves string categories
- Corrected degenerate Welch/ANOVA and multiple-testing behavior.
- Enforced valid sample-to-group and paired-study designs.
- Honored metadata missing-value masks.
- Prevented artifact-slot collisions from custom labels.
- Added actual selected-value fingerprints, preventing stale statistical reuse after matrix changes.
- Preserved immutable historical retrieval through exact ArtifactRefs.
- Fixed NumPy sequence inputs.
- Tightened plot/result compatibility across selections, groups, samples, pairs, normalization, and realized values.
- Fixed shared-axis annotations, dark/custom themes, missing pairs, invalid subsets, infinite values, and invalid bracket heights.
- Removed the plotting-to-datastore dependency and retained bounded per-feature fetching.
Quality assessment
…hed NygenAnalytics#177) into stats_testing

# Conflicts:
#	docs/source/tutorials/pseudobulk_and_differential_expression.md
#	scarf/datastore/_operations/features.py
#	scarf/datastore/_plot_accessor.py
#	scarf/features/statistical.py
#	scarf/metadata/arguments.py
#	scarf/plotting/distribution.py
#	tests/test_fate_helpers.py
#	tests/test_operation_argument_contracts.py
#	tests/test_plotting_modernization.py
#	tests/test_statistical.py
#	tests/test_statistical_datastore.py
Reframe data_organization around checkpoints and revert boxes: Short
Refresher with Scanpy/Seurat equivalents, live-I restore demos with
update_key/reset_key/insert/drop, immutable-selection notes for
filter_cells, and a revert cheat-sheet. Add a crash course on
cell-specific stats testing to condition_comparisons, adapted to the
donor-level paired design. Refresh the executed cache for the
reworked data_organization page.
@saatvik-viniak saatvik-viniak changed the title docs: more pedagogic data_organization and condition_comparisons docs: more pedagogic tutorials for data_organization and condition_comparisons Sep 5, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants