# Evaluation The crate includes accuracy and performance evaluations. Below are instructions on how to run those. _Note that all of these should be run from the crate's root directory!_ ## Accuracy See [accuracy results](evaluation/accuracy.md) for the predefined configurations. The evaluation tool is available as a CLI to run experiments on predefined and custom configurations. 1. Ensure R and ImageMagick are installed with necssary packages: - Install R from [download](https://cran.r-project.org/) or using your platform's package manager. - Start `R` and install packages by executing `install.packages(c('dplyr', 'egg', 'ggplot2', 'readr', 'stringr'))`. - Install ImageMagick using the official [instructions](https://imagemagick.org/script/download.php). 2. Run the following in the `geo_filters` crate root to see the help text: script/accuracy -h For example, to compare the predefined diff count configurations, run: $ script/accuracy geo_diff_{7,13} +--------------------+ + Running Simulation + +--------------------+ Parameters: number of configs = 2 number of samples = 500 number of sets = 120 Running simulations: geo_diff_7 ... done (0.70 s) geo_diff_13 ... done (1.71 s) +--------------------+ + Writing Results + +--------------------+ done csv file = accuracy.csv ... Output: accuracy.pdf Open `accuracy.pdf` to see the results. To regenerate the plots in the repository, ensure ImageMagick is available, and run: script/generate-accuracy-plots _Note that this will always re-run the experiments!_ ## Performance See [performance results](evaluation/performance.md) for the predefined configurations. To reproduce the benchmark (note that this may take quite some time), run: script/performance Open `../../target/criterion/report/index.html` to see the results. To regenerate the plots in the repository, ensure ImageMagick is available, and run: script/generate-performance-plots _Note that this will always re-run the benchmark!_