#moonbit-fish-stock

#What it provides

#Feature map

• Core types and validation — stock series, catch records, model parameters, typed errors, and simulation configuration.
• Math — finite-value-safe scalar approximations, statistics, linear algebra, optimization, and deterministic PRNG utilities.
• CPUE — raw/standardized CPUE, normalization, detrending, seasonal adjustment, abundance indices, and trend diagnostics.
• Surplus production — Schaefer, Pella–Tomlinson, and Fox dynamics, fitting, reference points, AIC/BIC, and residual diagnostics.
• Age-structured models — Beverton–Holt and Ricker recruitment, VPA, Baranov catch, Leslie-style transitions, SSB, and yield.
• Environment — anomaly, moving-average, lag-correlation, and climate-linked carrying-capacity helpers.
• Uncertainty and management — residual bootstrap, HCR/TAC calculations, closed-loop operating models, and MSE projections.
• Quality and indices — explicit series-quality reports, weighted survey-index aggregation, confidence intervals, and index comparison.
• Projection and performance — multi-year age projections, projection summaries, empirical MSE risk metrics, distribution summaries, stable ranking, and Markdown tables.
• Reporting — ASCII line charts and Markdown assessment reports.

#Package layout

moonbit-fish-stock/
├── lib/
│   ├── types/          core records, errors, validation, series operations
│   ├── math/           statistics, linear algebra, optimization, PRNG
│   ├── cpue/           CPUE standardization and abundance-index helpers
│   ├── surplus/        surplus-production models and diagnostics
│   ├── age_structured/ recruitment, VPA, age transitions, SSB, yield
│   ├── environment/    environmental covariates and climate dynamics
│   ├── bootstrap/      residual resampling and uncertainty summaries
│   ├── indicators/     reference points and Kobe status trajectories
│   ├── management/     HCR and TAC recommendations
│   ├── mse/            operating model and closed-loop simulations
│   ├── quality/        non-mutating input-quality diagnostics
│   ├── indices/        weighted annual survey-index summaries
│   ├── projection/     multi-year age-structured projections
│   ├── performance/    scenario metrics and ranking
│   ├── benchmarking/   deterministic cross-package benchmark workload
│   └── report/         terminal charts and Markdown reports
└── cmd/
    ├── main/           end-to-end assessment demonstration
    └── benchmark/      reproducible benchmark entry point

#Requirements

• MoonBit stable CLI. The current local verification uses moon 0.1.20260819 and moonc 0.10.9+6e6c44045.
• Git for cloning and version control.

#Quick start

git clone https://github.com/lfbnntr/moonbit-fish-stock.git
cd moonbit-fish-stock

# Type-check every supported backend and run all tests.
moon check --target all
moon test --target all

# Run the end-to-end assessment example.
moon run cmd/main

# Run the deterministic cross-package benchmark.
moon run cmd/benchmark

moon check --deny-warn
moon test --deny-warn
moon fmt --check
moon info

#Library example

import {
  "lfbnntr/moonbit-fish-stock/lib/types"
  "lfbnntr/moonbit-fish-stock/lib/quality"
  "lfbnntr/moonbit-fish-stock/lib/indices"
  "lfbnntr/moonbit-fish-stock/lib/projection"
  "lfbnntr/moonbit-fish-stock/lib/age_structured"
}

fn assess(series : @types.StockSeries) raise @types.ErrorType {
  let quality = @quality.inspect_stock_series(
    series,
    @quality.QualityConfig::strict(),
  )
  @quality.require_quality(quality)

  let index = @indices.aggregate_cpue_indices(series, 1.96)
  println("annual index years=" + index.length().to_string())

  let ages = @age_structured.AgeStructureConfig::new(
    3, 0.2, [1.0, 2.0, 3.0], [0.0, 0.5, 1.0], [0.0, 0.5, 1.0],
  )
  let path = @projection.project_age_structure(
    [100.0, 80.0, 50.0],
    @projection.AgeProjectionConfig::new(2025, 10, 0.1, 0.01, true),
    ages,
  )
  let summary = @projection.summarize_projection(path)
  println("projection depletion=" + summary.depletion.to_string())
}

#Reproducible benchmark

#Tests and continuous integration

moon test lib/quality --deny-warn
moon test lib/indices --deny-warn
moon test lib/projection --deny-warn
moon test lib/performance --deny-warn

#Modeling scope and limitations

• Statistical intervals in lib/indices use an explicit normal approximation around weighted annual estimates.
• MSE probabilities in lib/performance are empirical fractions of simulated trajectories, not guarantees about future stock status.
• Projection behavior is determined by supplied biological parameters and recruitment/fishing paths; users remain responsible for choosing defensible assumptions.
• The repository does not silently repair missing years, duplicate records, negative values, or non-finite observations. Inspect with lib/quality before fitting or projecting.

#License