Skip to content

[heatmap-cohort-retention] Cohort Retention Heatmap #4570

Closed
Closed
[heatmap-cohort-retention] Cohort Retention Heatmap#4570
Labels
approvedApproved for implementationimpl:altair:doneAltair implementation mergedimpl:bokeh:donebokeh implementation mergedimpl:chartjs:donechartjs implementation mergedimpl:d3:doned3 implementation mergedimpl:echarts:doneecharts implementation mergedimpl:ggplot2:doneggplot2 implementation mergedimpl:highcharts:donehighcharts implementation mergedimpl:letsplot:doneletsplot implementation mergedimpl:makie:donemakie implementation mergedimpl:matplotlib:doneMatplotlib implementation mergedimpl:muix:donemuix implementation mergedimpl:plotly:doneplotly implementation mergedimpl:plotnine:doneplotnine implementation mergedimpl:pygal:donepygal implementation mergedimpl:seaborn:doneseaborn implementation mergedspec-readySpecification merged to mainwatchdog:retried-plotnineWatchdog retried plotnine once

Description

@MarkusNeusinger
MarkusNeusinger
opened on Mar 6, 2026

Description

A triangular heatmap showing user retention rates across cohorts (rows) and time periods (columns). Each row represents a signup cohort, each column represents periods since signup, with cell color indicating retention percentage.

Key Visual Elements

  • Cohort labels on y-axis (e.g., "Jan 2024", "Feb 2024")
  • Periods since signup on x-axis (Week 0, Week 1, ...)
  • Triangular shape (later cohorts have fewer data points)
  • Color intensity: dark (high retention) to light (low retention)
  • Percentage values in cells
  • Week 0 always 100%

Applications

  • SaaS/product analytics: measuring user stickiness
  • Mobile apps: tracking engagement decay
  • Gaming: player retention analysis
  • Subscription businesses: churn pattern identification

Domain

Product Analytics / Marketing

Activity

  1. MarkusNeusinger
    added
    spec-requestNew specification request
    on Mar 6, 2026
  2. github-project-automation
    added this to Plot Backlogon Mar 6, 2026
  3. github-project-automation
    moved this to Todo in Plot Backlogon Mar 6, 2026

  4. claude commented on Mar 6, 2026

    @claude
    claudeBot
    on Mar 6, 2026 – with Claude
    Contributor

    ✅ Spec ID: heatmap-cohort-retention

    Rationale: Combines the heatmap plot type with the cohort-retention variant, clearly describing a triangular retention heatmap used in product analytics — distinct from existing heatmap specs (annotated, calendar, correlation, clustered, etc.).

    Add label approved to begin generation.

  5. github-actions
    added a commit that references this issue on Mar 6, 2026

    spec: add heatmap-cohort-retention specification

    16dc5c0
  6. claude
    changed the title [-]Cohort Retention Heatmap[/-] [+][heatmap-cohort-retention] Cohort Retention Heatmap[/+] on Mar 6, 2026

  7. github-actions commented on Mar 6, 2026

    @github-actions
    github-actionsBot
    on Mar 6, 2026 – with GitHub Actions
    Contributor

    ✅ Specification Ready: heatmap-cohort-retention

    PR: #4598

    specification.md

    heatmap-cohort-retention: Cohort Retention Heatmap

    Description

    A triangular heatmap displaying user retention rates across signup cohorts and time periods. Each row represents a cohort (e.g., users who signed up in a specific month), each column represents periods since signup, and cell color intensity indicates the retention percentage. The triangular shape naturally emerges because more recent cohorts have fewer elapsed periods. This visualization reveals retention trends, highlights churn patterns, and enables comparison of cohort quality over time.

    Applications

    • SaaS product analytics: tracking weekly or monthly user retention to measure feature stickiness and identify engagement drops
    • Mobile app growth: comparing retention curves across acquisition channels or app versions to optimize onboarding
    • Subscription business monitoring: identifying seasonal churn patterns and evaluating the impact of retention interventions
    • Gaming analytics: measuring player return rates across cohorts to assess content update effectiveness

    Data

    • cohort (string) - Cohort label representing the signup period (e.g., "Jan 2024", "Feb 2024")
    • period (integer) - Number of periods since signup (0, 1, 2, ...), where period 0 is the signup period
    • retention_rate (float) - Percentage of users retained, ranging from 0 to 100; period 0 is always 100%
    • cohort_size (integer) - Number of users in each cohort (displayed alongside cohort labels)
    • Size: 8-12 cohorts with 8-12 periods each
    • Example: Monthly signup cohorts from Jan 2024 to Oct 2024, with weekly retention percentages

    Notes

    • Period 0 (signup period) should always show 100% retention for every cohort
    • The heatmap should have a triangular shape: the first cohort has the most columns, each subsequent cohort has one fewer
    • Use a sequential colormap from light (low retention) to dark (high retention), such as a green or blue gradient
    • Display the retention percentage as text inside each cell
    • Show cohort size (number of users) next to each cohort label on the y-axis
    • X-axis labels should read "Week 0", "Week 1", etc. (or "Month 0", "Month 1" depending on the period granularity)
    • Consider adding a color bar legend to indicate the retention scale

    Next: Add approved label to merge this specification to main.

    🤖 spec-create workflow

  8. MarkusNeusinger
    added
    approvedApproved for implementation
    and removed
    approvedApproved for implementation
    on Mar 6, 2026
  9. github-actions
    added a commit that references this issue on Mar 6, 2026

    spec: add heatmap-cohort-retention specification (#4598)

    0c5692a
  10. github-actions
    added
    spec-readySpecification merged to main
    and removed
    spec-requestNew specification request
    on Mar 6, 2026

  11. github-actions commented on Mar 6, 2026

    @github-actions
    github-actionsBot
    on Mar 6, 2026 – with GitHub Actions
    Contributor

    🚀 Specification Merged!

    Specification: heatmap-cohort-retention
    PR: #4598 (merged)

    The specification is now in main and synced to PostgreSQL.

    Next: Add generate:{library} labels to start generating implementations.

    Available libraries: matplotlib, seaborn, plotly, bokeh, altair, plotnine, pygal, highcharts, letsplot

    Or use workflow dispatch for bulk operations.

    🤖 spec-create workflow

  12. MarkusNeusinger
    added
    generate:matplotlibTrigger matplotlib implementation
    on Mar 16, 2026

  13. 152 remaining items

  14. github-actions commented on Jul 5, 2026

    @github-actions
    github-actionsBot
    on Jul 5, 2026 – with GitHub Actions
    Contributor

    ⚠️ plotnine Generation Failed (Attempt 2/3)

    Attempt 2 failed. Automatically retrying...

    🤖 impl-generate

  15. github-actions
    added
    impl:plotnine:failedplotnine max retries exhausted
    and removed
    impl:plotnine:pendingplotnine generation in progress
    on Jul 5, 2026

  16. github-actions commented on Jul 5, 2026

    @github-actions
    github-actionsBot
    on Jul 5, 2026 – with GitHub Actions
    Contributor

    ❌ plotnine Failed (Attempt 3/3)

    The plotnine implementation for heatmap-cohort-retention failed after 3 attempts.

    Reason: Claude Code failed to create the implementation file.

    To retry manually:

    gh workflow run impl-generate.yml -f specification_id=heatmap-cohort-retention -f library=plotnine -f issue_number=4570 -f model=sonnet

    🤖 impl-generate

  17. github-actions
    added
    watchdog:retried-plotnineWatchdog retried plotnine once
    on Jul 6, 2026
  18. MarkusNeusinger
    added
    generate:plotnineTrigger plotnine implementation
    on Aug 17, 2026
  19. github-actions
    added
    impl:plotnine:pendingplotnine generation in progress
    on Aug 17, 2026

  20. github-actions commented on Aug 17, 2026

    @github-actions
    github-actionsBot
    on Aug 17, 2026 – with GitHub Actions
    Contributor

    🎨 python/plotnine Preview

    Light Dark
    plotnine light plotnine dark

    PR: #10309

    🤖 impl-generate

  21. MarkusNeusinger
    added a commit that references this issue on Aug 17, 2026

    feat(plotnine): implement heatmap-cohort-retention (#10309)

    79e8bb0
  22. github-actions
    removed
    generate:plotnineTrigger plotnine implementation
    impl:plotnine:pendingplotnine generation in progress
    impl:plotnine:failedplotnine max retries exhausted
    on Aug 17, 2026

  23. github-actions commented on Aug 17, 2026

    @github-actions
    github-actionsBot
    on Aug 17, 2026 – with GitHub Actions
    Contributor

    ✅ plotnine Complete

    plotnine implementation for heatmap-cohort-retention has been merged to main.

    PR: #10309
    Quality Score: 95/100

    Preview: light · dark

    🤖 impl-merge

  24. github-actions commented on Aug 17, 2026

    @github-actions
    github-actionsBot
    on Aug 17, 2026 – with GitHub Actions
    Contributor

    🎉 All Implementations Complete!

    All 15 library implementations for heatmap-cohort-retention have been successfully merged.

    Library Status
    altair ✅
    bokeh ✅
    chartjs ✅
    d3 ✅
    echarts ✅
    ggplot2 ✅
    highcharts ✅
    letsplot ✅
    makie ✅
    matplotlib ✅
    muix ✅
    plotly ✅
    plotnine ✅
    pygal ✅
    seaborn ✅

    🤖 impl-merge

  25. github-actions
    closed this as completedon Aug 17, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    approvedApproved for implementationimpl:altair:doneAltair implementation mergedimpl:bokeh:donebokeh implementation mergedimpl:chartjs:donechartjs implementation mergedimpl:d3:doned3 implementation mergedimpl:echarts:doneecharts implementation mergedimpl:ggplot2:doneggplot2 implementation mergedimpl:highcharts:donehighcharts implementation mergedimpl:letsplot:doneletsplot implementation mergedimpl:makie:donemakie implementation mergedimpl:matplotlib:doneMatplotlib implementation mergedimpl:muix:donemuix implementation mergedimpl:plotly:doneplotly implementation mergedimpl:plotnine:doneplotnine implementation mergedimpl:pygal:donepygal implementation mergedimpl:seaborn:doneseaborn implementation mergedspec-readySpecification merged to mainwatchdog:retried-plotnineWatchdog retried plotnine once

    Projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions