Skip to content

CLAUDE.md

Latest commit

 

History

History
102 lines (75 loc) · 2.77 KB

CLAUDE.md

File metadata and controls

102 lines (75 loc) · 2.77 KB

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Package Overview

hypothesize is an R package providing a consistent API for hypothesis testing, designed around principles from Structure and Interpretation of Computer Programs (SICP):

  1. Data Abstraction: Tests are S3 objects with accessor functions (pval(), test_stat(), dof())
  2. Closure Property: Combining tests yields tests (fisher_combine())
  3. Higher-Order Functions: Transform tests into new tests (adjust_pval())

Development Commands

# Generate documentation from roxygen2 comments
devtools::document()

# Run tests (83 tests covering all functions)
devtools::test()

# Run tests with coverage
covr::package_coverage()

# Check package (R CMD check)
devtools::check()

# Build vignettes
devtools::build_vignettes()

# Install from source
devtools::install()

Architecture

Core Design Pattern

hypothesis_test (base constructor)
       │
       ├── z_test (simplest primitive)
       ├── wald_test (general parameters)
       ├── lrt (model comparison)
       ├── fisher_combined_test (from fisher_combine)
       └── adjusted_test (from adjust_pval)

Generic Methods (the API)

All hypothesis test objects implement:

  • pval(x) — extract p-value
  • test_stat(x) — extract test statistic
  • dof(x) — extract degrees of freedom
  • is_significant_at(x, alpha) — check significance

Additional methods for specific types:

  • confint(x) — extract confidence interval (wald_test, z_test)

Key Functions

Function Purpose SICP Principle
hypothesis_test() Base constructor Data abstraction
z_test() One-sample z-test Primitive
wald_test() Test parameter = value Primitive
lrt() Compare nested models Primitive
fisher_combine() Combine p-values Closure property
adjust_pval() Multiple testing correction Higher-order function
confint() CI from test Duality

Extension Pattern

To add a new hypothesis test:

my_test <- function(data, null_value) {
  stat <- compute_statistic(data, null_value)
  p.value <- compute_pvalue(stat)
  hypothesis_test(
    stat = stat,
    p.value = p.value,
    dof = length(data) - 1,
    superclasses = "my_test",
    data = data,
    null_value = null_value
  )
}

The new test automatically inherits all generic methods.

Key Files

  • R/hypothesize.R — All package code (~700 lines)
  • tests/testthat/test-hypothesis-tests.R — 83 tests
  • vignettes/introduction.Rmd — Pedagogical vignette

Package Dependencies

  • Imports: stats (pchisq, pnorm, qnorm, p.adjust)
  • Suggests: testthat, knitr, rmarkdown