This repository contains scripts that help to reproduce the results of the paper "Faster Block Tree Construction".
To clone this repository and build the benchmark please use the following commands.
git clone https://github.com/pasta-toolbox/block_tree_experiments --recursive
cd block_tree_experiments
cmake -DCMAKE_BUILD_TYPE=Release -B build
cmake --build build -j
This will compile all code needed for the benchmark in the build folder, which is used in the run_benchmarks.sh script.
Next, we can run the script.
./run_benchmarks.sh <path to Pizza&Chili repetitive corpus>
This script will copy the template folder and rename it to results_CURRENTDATE, where CURRENTDATE can be decoded using the date command (date --date @CURRENTDATE).
Within the folder, you find a LaTex-Documents that contains all layout and styling information needed to recreate the plots.
To make the plots based on the newly created data, we rely on sqlplot-tools.
This tool reads the newly created results files and passes all values directly into the LaTeX files.
Please refer to sqlplot-tools' website for information on how to install it.
If you have installed sqlplot-tools simply run sqlplot-tools <file_name.tex> and compile <file_name.tex> to see the final results.
Warning: if you are using a different text corpus for the experiments, you have to adapt the plots. Otherwise, no data will be shown, as the name of the inputs are used to populate the plots.