Skip to content

plan.md

Latest commit

 

History

History
84 lines (66 loc) · 4.63 KB

plan.md

File metadata and controls

84 lines (66 loc) · 4.63 KB
Raw

Implementation plan

The first milestone is a reproducible, validated headline life-expectancy graph embedded in WordPress. This checklist tracks completed setup and proposed implementation work. See architecture.md for the current system, proposed layers, and definition choices.

1. Agree on the chart and audit sources

  • Confirm the written specification versus the mockup's different series.
  • Define country/territory eligibility, the annual population threshold, frontier ties, global median, and income-group aggregation and membership.
  • Audit UN WPP and OWID artifacts for female/male life expectancy at birth and 65, annual population, income groups, units, and historical coverage.
  • Choose and record source releases, citations, licences, and permitted uses.
  • Assemble a small verified sample spanning countries, years, sexes, and ages.
  • Record missing-data rules and keep historical estimates separate from projections.

Done when the series definitions and source mapping are documented and sample values can be traced back to the source.

2. Make the pipeline reliable locally

  • Derive the repository root independently of .env discovery.
  • Always build in ignored output/; remove the local/non-local distinction.
  • Prepare snapshot provenance for review and require matching committed records before publication.
  • Separate publication into an explicit command with upload recovery.
  • Document Python setup and record resolved dependency versions for Python 3.10/Linux.
  • Optional: restore a selected snapshot from storage, verify its checksum, and rebuild without fetching newer upstream data. Not required for initial local development.
  • Make snapshot publication retries independent of importer completion.
  • Reject or explicitly align mismatched multipart CSV headers.
  • Track metadata-only upstream changes as well as CSV changes.
  • Add focused regression tests for upload failure/retry and publication validation.
  • Add focused regression tests for CSV schema mismatch.

Done when a local build requires no cloud credentials and failed publication can be retried without deleting or re-downloading a valid snapshot.

3. Build standardised data and indicators

  • Implement the selected source adapter and explicit sex/age/measure schema.
  • Add stable location IDs, population references, and versioned income classifications.
  • Validate dimension uniqueness, required units, join coverage, and missing values.
  • Calculate country, frontier, median, and income-group series.
  • Preserve frontier winners, input snapshot IDs, and calculation code version.
  • Test the one-million boundary, ties, missing years, aggregate exclusion, sex/age isolation, and chosen aggregation rules with small known examples.
  • Cross-check representative output values against source data.

Done when the highlighted series are independently checked and reproducible from recorded inputs.

4. Build the headline graph

  • Define a chart manifest with a stable ID, series metadata, sources, and defaults.
  • Render grey country lines and the four highlighted series with clear labels.
  • Add a female-default, keyboard-accessible sex toggle and informative tooltips.
  • Include methodology, coverage, source attribution, and a data download.
  • Add loading, empty-data, and error states.
  • Check mobile layout, contrast, keyboard access, and an accessible text/table alternative.
  • Validate the at-birth chart, then add the age-65 view with remaining-years units.

Done when the chart matches the agreed definitions, remains readable on mobile, and its controls and values have been checked.

5. Publish and embed

  • Package immutable data, metadata, and rendering assets into a release.
  • Validate uploaded artifacts before updating the public release pointer.
  • Check content types, caching, cross-origin access where needed, and embed behavior.
  • Embed in WordPress and verify sizing, interaction, and download links.
  • Exercise rollback and a rebuild from the recorded snapshot and code revision.
  • Add CI checks and document the update/review/publication workflow.

Done when the live embed works and the release can be rebuilt or rolled back.

Later milestones

  • Annual changes and ten-year improvement rates, with explicit window rules.
  • Country-to-frontier gaps and selected-country comparisons.
  • Country dashboards consuming the shared indicators.
  • Validate IHME provenance before adding healthy-life-expectancy measures.
  • Add life-table inputs for lifespan percentiles and disparity measures.