run-relnet-asp.py: script to run RelNet-ASPadd_chain_formula.py: script to compute chain ASP programmolise.pl: an example graph instance
The Benchmarks are available here.
The ApproxASP is publicly available here: ApproxASP. One binary of ApproxASP is given in the current directory.
Please check whether add_chain_formula.py, approxasp, molise.pl exist in your current directory, and approxasp is executable (chmod +x)
The input graph is molise.pl (LP format). The command to compute network reliability of molise.pl for
python run-relnet-asp.py -i molise.pl -k 1 -m 3
Please check whether add_chain_formula.py, approxasp, molise.pl exist in your current directory, and approxasp is executable (chmod +x)
First compute the chain formula of molise.pl for edge probability
python add_chain_formula.py -i molise.pl -k 1 -m 3
After successful execution, the command will show the following output:
Number of new rules added: 250
The multiplication factor: 375
We need to value of The multiplication factor: to compute the network reliability. More specifically, we divide the ASP count by add_chain_formula.py will generate two files to run ASP counter: chain formula augmented ASP program (chain_molise.pl and independent support IS_chain_molise.pl. Independent support is useful for counting efficiently.
Now let run an ASP counter using the following command:
./approxasp --sparse --conf 0.35 --useind IS_chain_molise.pl --asp chain_molise.pl
The command will take few seconds to run and finally it shows the approximate count in line: After the iteration, the (median) number of solution: 50 * 2 ^ 356
So, the reliability is
If you face any issue, create a new issue or email to [email protected].
@inproceedings{KM2023,
title={A Fast and Accurate ASP Counting Based Network Reliability Estimator.},
author={Kabir, Mohimenul and Meel, Kuldeep S},
booktitle={LPAR},
volume={94},
pages={270--287},
year={2023}
}