ACRO: Tools for the Automatic Checking of Research Outputs

This repository maintains the ACRO R package, which is an interface to the Python ACRO package.

ACRO (Automatic Checking of Research Outputs) is an open source tool for automating the statistical disclosure control (SDC) of research outputs. ACRO assists researchers and output checkers by distinguishing between research output that is safe to publish, output that requires further analysis, and output that cannot be published because of substantial disclosure risk.

It does this by providing a light-weight 'skin' that sits over well-known analysis tools, in a variety of languages researchers might use. This adds functionality to:

• identify potentially disclosive outputs against a range of commonly used disclosure tests;
• suppress outputs where required;
• report reasons for suppression;
• produce simple summary documents TRE staff can use to streamline their workflow.

Installation

Install the acro package from CRAN as follows:

install.packages("acro")

Notes for Python 3.12

ACRO currently depends on an older version of Pandas (~1.5.0) for which no pre-compiled wheels are available within pip for Python 3.12. Therefore, in this scenario, Pandas must be built from source. This requires the installation of a C++ compiler before pip installing acro. Earlier Python versions do not have this issue.

For Windows, Microsoft Visual Studio and the C++ build tools will likely need to be installed first.

If you are unable to install the C++ tools, but are able to install multiple versions of Python, you can use the following method to explicitly create the Python virtual environment with your desired python.exe.

First identify the location of the python.exe needed, which may be in C:\Program Files\ or in your local C:\Users\username\AppData\Local\Programs\Python\.

Then from R, use reticulate to create the virtual environment and install ACRO; see example below. Change the path provided as a version to your Python path (note that forward slashes are required in the path).

>>> library("reticulate")
>>> reticulate::virtualenv_create(envname = "r-acro-0.4.6", version = "C:/Users/username/AppData/Local/Programs/Python/Python310/python.exe", force = TRUE, packages = NULL)
>>> reticulate::py_install("acro", envname = "r-acro-0.4.6")

Finally, the R ACRO library can be used as normal:

>>> library("acro")
>>> acro_init(suppress = TRUE)

Usage

Before using any function from the package, an acro object should be initialised using the following R code:

>>> library("acro")
>>> acro_init()

Then the functions can be called.

Acknowledgement

This work was funded by UK Research and Innovation under Grant Numbers MC_PC_21033 and MC_PC_23006 as part of Phase 1 of the Data and Analytics Research Environments UK (DARE UK) programme, delivered in partnership with Health Data Research UK (HDR UK) and Administrative Data Research UK (ADR UK). The specific projects were Semi-Automatic checking of Research Outputs (SACRO; MC_PC_23006) and Guidelines and Resources for AI Model Access from Trusted Research environments (GRAIMATTER; MC_PC_21033). This project has also been supported by MRC and EPSRC [grant number MR/S010351/1].