This repository maintains the ACRO R package, which is an interfaceto the PythonACROpackage.
ACRO is a free and open source tool that supports the semi-automatedchecking of research outputs (SACRO) for privacy disclosure withinsecure data environments. SACRO is a framework that appliesbest-practice principles-basedstatisticaldisclosure control (SDC) techniques on-the-fly as researchersconduct their analysis. SACRO is designed to assist human checkersrather than seeking to replace them as with current automatedrules-based approaches.
The ACRO package is a lightweight Python tool that sits overwell-known analysis tools that produce outputs such as tables, plots,and statistical models. This package adds functionality to:
- automatically identify potentially disclosive outputs against arange of commonly used disclosure tests;
- apply optional disclosure mitigation strategies as requested;
- report reasons for applying SDC;
- and produce simple summary documents trusted research environmentstaff can use to streamline their workflow and maintain auditablerecords.
This creates an explicit change in the dynamics so that SDC issomething done with researchers rather than to them, and enables moreefficient communication with checkers.
A graphical user interface (SACRO-Viewer) supportshuman checkers by displaying the requested output and results of thechecks in an immediately accessible format, highlighting identifiedissues, potential mitigation options, and tracking decisions made.
Additional programming languages such as this R package are supportedby providing front-end packages that interface with the core ACRO Pythonback-end.
Installation
Prerequisite: you must havePython pre-installed.
Install theacro package from CRAN as follows:
install.packages("acro" )If you are having problems installing the package, please see themore detailedinstallation guide.
Usage
Before using any function from the package, an acro object should beinitialised using the following R code:
>>>library("acro")>>>acro_init(suppress =TRUE )Try Online with MyBinder
Try an example notebook inRStudioonline on MyBinder.org.
Select theexample-notebook.Rmd in the bottom-right paneafter loading.
Documentation
The github-pages contains pre-builtdocumentation.
Additionally, see ourpaper describing theSACRO framework to learn about its principles-based SDC methodologyand usage.
Acknowledgement
This work was funded by UK Research and Innovation under GrantNumbers MC_PC_21033 and MC_PC_23006 as part of Phase 1 of the Data andAnalytics Research Environments UK (DARE UK) programme, delivered inpartnership with Health Data Research UK (HDR UK) and AdministrativeData Research UK (ADR UK). The specific projects were Semi-Automaticchecking of Research Outputs (SACRO; MC_PC_23006) and Guidelines andResources for AI Model Access from Trusted Research environments(GRAIMATTER; MC_PC_21033). This project has also been supported by MRCand EPSRC [grant number MR/S010351/1].
[8]ページ先頭