Encode Bayesian Networks into Conjunctive Normal Form and write it to output file in DIMACS format. (STAND ALONE VERSION)
The program takes a HUGIN encoded Bayesian Network as input and encodes it into Conjunctive Normal Form with various optimizations, e.g., prime implicates, context-specific independence, determinism, etc.
./bn-to-cnf -i <HUGIN FILE> [option] [...]
| Option | Optimization |
|---|---|
| -p | Partition cnf per CPT |
| -c | Constraints are suppressed |
| -e | Equal probabilities are encoded |
| -d | Determinism are encoded |
| -a | Apply boolean symplification |
| -b | Boolean variables are not mapped |
| -q | Quine-McCluskey (QM) |
| -l <limit> | Limit problem size for QM |
| Option | Other |
|---|---|
| -i <filename> | Input (HUGIN .net file) |
| -w | Write CNF in DIMACS format to file |
| -s | Show stats |
| -h | Help |
Just type:
make