The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

Version 1.1.2

• Renamed the mec_blocking() trade-off argument from alpha to rho to avoid confusion with the hurdle Gamma shape parameter. The old alpha argument remains accepted as an undocumented compatibility alias.
• Added join_records() to join files after the MEC record linkage procedure.

Version 1.1.1

• Improved mec_blocking() by using inverted unsupervised MEC.
• Added the trade-off argument in mec_blocking() for controlling the FLR-MMR trade-off.

Version 1.1.0

• Added mec_blocking() for blocked unsupervised MEC with pooled training and blockwise prediction using the blocking package.
• Added support for creating comparison vectors on a supplied table of record pairs through the pairs argument in comparison_vectors().
• Added census and cis example datasets for larger record linkage examples.
• Added a vignette showing MEC with blocking on the cis and census datasets.
• Added optional progress messages via the verbose argument in mec(), train_rec_lin(), predict.rec_lin_model(), and mec_blocking().
• Improved validation of supplied match and pair tables, including clearer checks for row indices, duplicate pairs, missing values, and non-finite comparison values.
• Improved print methods for linkage results, including consistent percentage formatting for error rates.

Version 1.0.1

• Fixed CRAN errors.

Version 1.0.0

• Implemented comparison functions abs_distance() and jarowinkler_complement().
• Added support for comparing two datasets using comparison functions.
• Added support for training a supervised record linkage model using probability or density ratio estimation, based on the following methods: "binary", "continuous_parametric", and "continuous_nonparametric".
• Added support for creating a supervised record linkage model using a custom machine learning (ML) classifier.
• Added support for predicting matches based on a record linkage model.
• Added the unsupervised maximum entropy classification (MEC) algorithm for record linkage. Supported methods are: "binary", "continuous_parametric", "continuous_nonparametric", and "hit_miss".
• Added support for creating the predicted set of matches based on: its estimated size, a target false link rate (FLR) or a target missing match rate (MMR).
• Implemented S3 methods for printing.
• Added support for evaluation when true matches are known.
• Added documentation and examples.

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.