Additional Material for the paper A Self-Adaptive Digital Twin Architecture for Dynamic Resource Management

Additional Material for the paper A Self-Adaptive Digital Twin Architecture for Dynamic Resource Management The repository contains additional material for the SEAMS 2026 conference paper.The material includes: The source code of the prototype implementation of the approach presented in the paper in the src folder. The evaluation data and results in the data folder. To process the data and reproduce the graphs presented in the paper, the following Python packages are required: matplotlib numpy the requirements.txt file can be used to install the required packages using pip: pip install -r requirements.txt Data evaluation The `data` folder contains the evaluation data and results for the experiments presented in the paper, alongside the scripts used to generate the plots.To execute the commands below, navigate to the `data` folder in your terminal cd data To reproduce the graph in Figure 9 in the paper, run the following command in the data folder: python3 plot_single.py sim_output_no_peaks/normal_40_15_1_20_True To reproduce the graph in Figure 10 in the paper, run the following command in the `data` folder: python3 plot_multi.py --mean 40 --std 15 --mode normal --iterations 5 --time_steps 20 --adaptive --base sim_output_peaks The results will be saved in the two folders, sim_output_no_peaks and sim_output_peaks, respectively. To generate the time statistics (with the different times captured) from the results, run the following commands in the data folder: python3 plot_time_multi.py --iterations 10 --mean 40 --std 15 --time_steps 20 --adaptive --base sim_output_no_peaks for the results without peak adaptation, and python3 plot_time_multi.py --iterations 10 --mean 40 --std 15 --time_steps 20 --adaptive --base sim_output_peaks for the results with peak adaptation.The parameters passed to the scripts are as follows: --mean: the lambda of the distribution used to generate the sensor data. --std: the standard deviation. While not used in the Poisson distribution, it still serves to identify the correct folder names. --mode: the distribution mode used to generate the distribution with different characteristics. We used `normal` in the experiments, and the parameter is used to identify the correct folder names. --iterations: the number of iterations used in the experiments. --time_steps: the number of time steps used in the experiments. --adaptive: flag to indicate that the adaptive strategy was used. --base: the base folder where the results are stored. Source code The src folder contains the source code of the prototype implementation of the approach presented in the paper. The README in the folder further shows how to install the system and run tests.