Repository navigation
refactor: remove the matplotlib plotting layer - #512
Merged
Merged
Conversation
Contributor
📊 Dashboard previewThe dashboard was built for this PR. ➡️ Download
|
exploreriii
force-pushed
the
refactor/remove-plotting
branch
from
October 3, 2026 17:31
3fdf880 to
1d01dc0
Compare
exploreriii
marked this pull request as draft
October 3, 2026 17:35
exploreriii
force-pushed
the
refactor/remove-plotting
branch
2 times, most recently
from
October 3, 2026 17:47
262adc5 to
4dd56dd
Compare
Signed-off-by: exploreriii <133720349+exploreriii@users.noreply.github.com>
…erged HIP work Restores the gate the retired renderer applied: zero rows stay in the dataset, but a chart in which every value is zero is not listed. The rule is now a declarative source option, hide_when_all_zero, so it lives in the spec rather than in a pipeline. Signed-off-by: exploreriii <133720349+exploreriii@users.noreply.github.com>
…c cannot reach 28 of 35 entries were already listed through chart sources or table sections, which let the orphan check cover an output after the spec stopped reading it. A guard test keeps the list from re-growing. Signed-off-by: exploreriii <133720349+exploreriii@users.noreply.github.com>
Signed-off-by: exploreriii <133720349+exploreriii@users.noreply.github.com>
427 errors across 50 files, almost all pandas type-inference unions that pyright cannot resolve without stubs; the report-only step produced noise rather than signal. Ruff remains the Python gate. Signed-off-by: exploreriii <133720349+exploreriii@users.noreply.github.com>
analysis is documented as I/O-free, but analysis.affiliation read data/affiliations.yaml itself. AFFILIATIONS_PATH, load_affiliations and load_manual_logins now live in data_sources.affiliations; the analysis module takes the mapping as data, as every other transform does. Signed-off-by: exploreriii <133720349+exploreriii@users.noreply.github.com>
The lint workflow checks .clusterfuzzlite/requirements.txt against uv.lock (uv export); dropping matplotlib changed the lock without regenerating it. Signed-off-by: exploreriii <133720349+exploreriii@users.noreply.github.com>
exploreriii
force-pushed
the
refactor/remove-plotting
branch
from
October 3, 2026 17:51
4dd56dd to
f9a0e87
Compare
MonaaEid
marked this pull request as ready for review
October 3, 2026 22:35
MonaaEid
approved these changes
Oct 3, 2026
danielmarv
approved these changes
Oct 4, 2026
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.
Description:
edit: sorry for the spam
Since #490 the dashboard has drawn every chart from the JSON chart documents the data API emits from pipeline CSVs. The PNGs had no reader left, but the Python side still treated the PNG as the primary key for a chart. This PR removes that plumbing and the plotting layer with it.
src/hiero_analytics/plotting/and its tests,export/save.plot_and_save, the provenance PNG footer, and every styling constant only a rendered image needed. Dropmatplotlib(and transitively pillow, contourpy, kiwisolver, fonttools, cycler, pyparsing).networkxstays: the network layout moves toanalysis/network_layout.pyand ships as data.files→variants,interactive_sources→sources.FULL_ROW_CHARTSdeclares the full-row layout that was measured from image dimensions before.file,width,height;interactiveis always present.pipelines/affiliation.pyintoconfig.charts+ a pureanalysis.affiliation.segment_colors, removing anexport → pipelinesimport.onboardingwritesonboarding_signal.csvandonboarding_efficiency.csv;contributor_churnwrites funnel, retention, transitions and tenure tables (andcontributor_progression.csvnow goes throughsave_dataframe, author as a column).contributor_profilesaddsmax_difficulty_distribution.csv.ChartVariant.fileremoved,interactiverequired, the no-data placeholder removed,/chartsdev proxy removed. CI stops copyingoutputs/chartsinto the Pages site. Docs describe provenance as sidecars + dashboard header/footer + snapshot.Behaviour changes worth knowing: