LoginSign Up
← Back to all sparks
R

reliaplotr

ANALYTICS
Velocity0.0

R package reliaplotr by paulgovan — release notes from GitHub.

github.com/paulgovan/reliaplotr ↗

The Weibull plotting package renamed itself, then handed its charts to AI assistants.

reliability-engineeringr-packageplotlymcpai-toolingdata-visualization
◆Current state
ReliaPlotR draws interactive reliability plots with plotly — probability plots, contour plots, Duane and reliability growth charts, accelerated life testing plots by stress level, mean cumulative function curves for repairable systems, and exposure plots. It was WeibullR.plotly until late 2025, and the rename tracked a real widening of scope rather than just a label change. The current release adds tidy extractors that turn fitted model objects into data frames, and an MCP server exposing five of its fit and plot functions as tools.
◆Where it's heading
The package has been following its analysis siblings function for function: as accelerated life testing and repairable systems modelling landed in the wider suite, the matching plot types appeared here, and when the growth-analysis package shipped an MCP server, this one followed two weeks later. The tidy extractors point the same way — a plotting package that can also return parameter estimates, goodness-of-fit metrics, and confidence bounds as tidy frames is one designed to be consumed programmatically, by a pipeline or an assistant, not only read on screen. Overlaying multiple model fits on a single plot has been a recurring request answered across several releases.
◆Prediction
Expect the tidy extractor and MCP tool surfaces to keep expanding together, since each new plot type in the suite now implies both a chart and a machine-readable version of what it shows.

◆Recent moves

  1. 2mo ago

    Tidy model extractors plus five plotting tools over MCP

    ⚡ SPARK

    Two changes that point the same direction: tidy_wblr(), tidy_rga() and tidy_alt() turn fitted objects into data frames of estimates, goodness-of-fit metrics and confidence bounds, and reliapltr_mcp_server() exposes fitting and plotting as MCP tools. Coming two weeks after the sibling growth-analysis package did the same, it confirms a suite-wide decision rather than one maintainer's experiment.

    View source ↗
  2. 2mo ago

    NHPP plots switch to the mean cumulative function

    plotly_nhpp() now draws the mean cumulative function rather than raw cumulative failures, which changes what the chart actually claims about a repairable system rather than how it looks. A rendering order fix for confidence bounds accompanies it — small corrections to plot types added only one release earlier.

    View source ↗
  3. 4mo ago

    Accelerated life testing and repairable systems plots added

    Five new plot types in one release — ALT probability plots with a line per stress level, life-stress relationship plots, mean cumulative function curves, NHPP growth plots, and exposure plots — mirroring the analysis capabilities landing in the sibling packages. Multiple model objects can now be overlaid on a single plot, the recurring ask for comparing competing fits.

    View source ↗
  4. 8mo ago

    Duane plots gain confidence bounds

    A confidence bounds option for the Duane plot and the removal of a results table judged to be clutter. Minor, but consistent with a package that keeps trading on-canvas detail for readable charts while moving the numbers to extractors.

    View source ↗
  5. 9mo ago

    ReliaPlotR v0.4.1

    A patch tag whose notes contain nothing but a changelog comparison link. What changed between this and the preceding release is not readable from the feed.

    View source ↗
  6. 10mo ago

    Renamed from WeibullR.plotly to ReliaPlotR

    The rename that declared the wider scope, dropping the Weibull-specific branding for reliability engineering generally. Read against what followed — ALT, repairable systems, and MCP tooling — it was an accurate signal rather than cosmetic, and the sibling tutorial package made the same move around the same time.

    View source ↗