SUMiT: Statistical Uncertainty Management Toolkit

The UWE Statistical Uncertainty Management Toolkit (SUMiT) is a software framework for co-ordinating the application of a range of techniques to large datasets in the production and publication of public statistics.

---

Contents

• build: empty folder to be used for building executables
• data: contains example data files for testing
• lib: contains third-party libraries
• python: contains example Python scripts
• sumit: contains SUMiT source code

---

Building

Requirements

• C++11 compliant compiler.
• Java JDK
• CMake (>= 3.12)
• Python (>= 3.6)

Compiler Options

• NATIVE_OPT=ON : Optimise for the native system architecture (CMake default = ON)

Ubuntu

$ sudo apt install python3 cmake
$ git clone --recurse-submodules https://github.com/AI-SDC/SUMiT.git
$ cd SUMiT/build
$ cmake -DCMAKE_BUILD_TYPE=Release ..
$ make

macOS

Notes: if brew is not an option, install CMake.

$ brew install cmake python
$ git clone --recurse-submodules https://github.com/AI-SDC/SUMiT.git
$ cd SUMiT/build
$ cmake -DCMAKE_BUILD_TYPE=Release ..
$ make

---

Running

Stand-alone

1. Start Server

$ java -jar sumit/sumit_server/cell_suppression_server.jar

2. Run Suppression Tool

$ ./sumit/cell_suppression_tool/cell_suppression_tool

Python

$ python3 example.py