The Pyramis Library: Efficient Numerical Evaluation of Hierarchical UML Statecharts applied to Stochastic Workflows

This repository is a companion page for the following submitted publication:

Laura Carnevali, Reinhard German, Leonardo Montecchi, Leonardo Scommegna, and Enrico Vicario. 2025. CThe Pyramis Library: Efficient Numerical Evaluation of Hierarchical UML Statecharts applied to Stochastic Workflows. (Currently under review).

This replication package includes all resources needed to reproduce our experimental results.

Prerequisites

• Java 23 or higher
• Maven 3.9.9 or higher

Setup Instructions

1. Clone this repository:

git clone git@github.com:oris-tool/pyramis-efficient-numerical-evaluation-rep-pkg.git
cd pyramis-efficient-numerical-evaluation-rep-pkg

2. Run the installation script:

./install.sh

Running the Experiments

Ground Truth Generation

Execute from the root directory:

mvn exec:java -Dexec.mainClass="it.unifi.hierarchical.epew25.GTGenerator"

This command will generate the Ground Truth (GT) in the "POSTP_GT" folder.

Important Notes:

• To skip processing time, use the pre-computed GT files in GT

Pyramis Analysis Experiment

mvn exec:java -Dexec.mainClass="it.unifi.hierarchical.epew25.PostProceedingsEpewExperiment" \
    -Dgt.path="[GT_PATH]" \
    -Dgt.runs="[GT_SIMULATION_RUNS]"

This command will perform the analysis with Pyramis for each configuration.

Default values for gt.path and gt.runs are set in order to work with the pre-calculated GT files. If you want to perform the analysis with the files in the GT directory, both parameters can be omitted.

If you want to perform the analysis with a different GT, please specify the path to the folder where the files are located (the [GT_PATH] parameter) and how many runs were used to obtain them (the [GT_SIMULATION_RUNS] parameter). By default, the generation script outlined in the previous step uses 1,000,000 simulation runs.

Pyramis Time Experiments

mvn exec:java -Dexec.mainClass="it.unifi.hierarchical.epew25.PostProceedingsTimeExperiment"

This command will calculate the mean time required to analyse each workflow configuration

Results Visualization

To visualize the experimental results, which will be stored in the results directory:

1. Install the required Python dependencies: From the scripts directory, run:

   pip install -r requirements.txt
   

2. Generate plots: Relies on the post-proceedings-single-plot.ipynb Python Notebook

File Structure

pyramis-efficient-numerical-evaluation-rep-pkg/
      |
      ├── install.sh                   # Installation script
      ├── src/                         # Source code 
      ├── GT/                          # Pre-calculated 1-million run Ground Truth CDFs
      └── scripts/                     # Python scripts for results visualization

Troubleshooting

If you encounter any issues during the replication process, please:

1. Ensure all prerequisites are correctly installed
2. Verify all required files are in the correct locations
3. Check the console output for error messages