studyframe 0.3.0

studyframe keeps observations with their collection roster, measurements, sampling and assignment declarations, weights, and source evidence. Analysis requests check those declarations before calling R survey or estimatr.

This experimental R release introduces the studyframe name, portable .studyframe bundles, supported Qualtrics CSV/QSF imports, structured analysis requirements and receipts, and an executable collection-to-inference vignette. It also fixes exact numeric joins, missing-code preservation, nonfinite metadata, undefined inference results, and empty filtered views.

The package, schema, and bundle format remain experimental. Earlier unreleased SuperDF APIs and bundles have no compatibility aliases. A filtered view retains and exports its complete attached collection roster. The package checks declared analysis requirements; it does not verify the truth of collection assertions or identification assumptions.

Install the R release and its analysis/bundle dependencies:

install.packages(c("digest", "jsonlite", "jsonvalidate", "survey", "estimatr", "arrow", "zip"))
install.packages(
  "https://github.com/gojiplus/superdf/releases/download/v0.3.0/studyframe_0.3.0.tar.gz",
  repos = NULL, type = "source"
)
library(studyframe)
vignette("collection-to-inference", package = "studyframe")

This release is distributed through GitHub; it has not been submitted to CRAN. The attached R source archive includes the compiled vignette. The repository also contains a Python bundle reader, but this release does not publish a Python package to PyPI.

Validation: 294 R assertions and 23 Python tests pass locally and in an independent review. R CMD check reports zero errors, warnings, or notes. The installed R source archive runs the synthetic collection, bundle handoff, experimental effect, and survey-mean examples.