coresynth 1.2.0

A review of the GSC, TASC, SI, SDID and MC (Matrix Completion) code against their papers and against gsynth/fect/MCPanel. Estimates from method = "gsc" with covariates, from method = "tasc", from method = "mc" with its default penalty, and from lambda_pen = "auto" change, as do the intervals and p-values of sdid_inference(), gsc_inference(), si_inference(), gsc_boot() and conformal_inference(). Each entry below says when and by how much.

Performance and computational improvements

Bug fixes

New features

Internal

coresynth 1.1.0

New features

coresynth 1.0.1

Bug fixes

Performance

coresynth 1.0.0

Breaking changes

New features

Bug fixes

coresynth 0.5.0

New features

Changed

coresynth 0.4.1

Changed

New features

Bug fixes

Performance

coresynth 0.4.0

Breaking changes

New features

Deprecated

Performance

Bug fixes

coresynth 0.3.0

New features

Improvements

Bug fixes

Internal

coresynth 0.2.4

Bug fixes

Input validation

Malformed panels previously produced confusing or misleading errors — for example, a 0-row data frame (from an upstream filtering bug) was misclassified as staggered adoption and failed with an unrelated “predictors not supported” error, and missing (id, time) cells surfaced as a raw eig_sym(): decomposition failed from the C++ layer. panel_to_matrices() (shared by all six estimators) and scm_design() now validate and error clearly on:

scm_fit() also now rejects non-numeric (factor/character) outcome or treatment columns and non-integer treatment values, instead of silently coercing them (as.integer(factor(...)) returns level codes, not the original values). “All cohort-level fits failed” errors (SCM/SDID/GSC/SI staggered paths) now point back to the preceding per-cohort warnings for diagnosis.

coresynth 0.2.3

Performance

coresynth 0.2.2

Bug fixes

coresynth 0.2.1

Bug fixes

New features

coresynth 0.2.0

New features

Minor improvements

coresynth 0.1.0

First public release.

Methods

Unified API

Performance

All core optimisations implemented in C++ via RcppArmadillo: 50–70x faster than the Synth package for typical panel sizes (N_co ≤ 30). src/inference.cpp placebo loops parallelised with OpenMP.