Building a solver for the puzzle Tents by encoding it as a SAT problem.
To start the GUI, simply run the "main.py" file. File "game.py" contains methods to read and solve game-fields. The read-me file additionally entails a full documentation on how to run the code and the individual functions per console.
All bonus tasks as well as the main task were fully integrated. We implemented and compared different encoding strategies with different SAT solvers. In particular, we implemented the binomial coding, binary coding, and sequential coding. We found the sequential coding is by far the fastest and most powerful coding strategy, especially with increasing field sizes. To fully appreciate the incredible speed we suggest to call/test the function per console to prevent the GUI bottleneck, although it still performs extremely well with the GUI, even on larger fields.
GUI is optimized for windows. Also tested and working on Linux (When you load the Stats Page you have to close the external PNG windows, as Linux displays is per default )
from Assignment_one.game import GameEncoderBinomial
# To read a game-field via a .txt file:
path = "...\\...\\..."
game_from_text = GameEncoderBinomial.from_text_file(path)
#To read a game-field directly via a game-id:
game_id = "...x..:....,...,...,.."
game_from_id = GameEncoderBinomial.from_game_id(game_id)
# To generate a random game-field:
game_random = GameEncoderBinomial.from_randomness(size=(8,8), tree_density=0.5).from Assignment_one.game import GameEncoderBinomial
# Create a game-field:
game = GameEncoderBinomial.from_...
# Print the unsolved game-field:
print(game.output_field())
# Solve the game-field:
game.solve_sat_problem()
# Print the solved game-field:
print(game.output_field())from Assignment_one.performance import print_encoding_details, get_encoding_details
# Create a game-field:
game = GameEncoderBinomial.from_...
# Print the number of variables, clauses and literals:
print_encoding_details(game)
# Get individual values
variables_n, literals_n, clauses_n = get_encoding_details(game)from Assignment_one.game import GameEncoderBinomial, as_DIMACS_CNF
# Create a game-field:
game = GameEncoderBinomial.from_...
# Get the CNF game encoding
cnf = game.get_cnf_solution()
# Transform and print it in DIMACS format:
print(as_DIMACS_CNF(cnf))from Assignment_one.performance import analyse_sat_solvers
g1 = GameEncoderBinomial.from_text_file("tent-inputs\\tents-10x10-t1.txt")
g2 = GameEncoderSequential.from_text_file("tent-inputs\\tents-10x10-t1.txt")
g3 = GameEncoderBinary.from_text_file("tent-inputs\\tents-10x10-t1.txt")
# Compare different SAT sovlers and store the result as a png. file.
analyse_sat_solvers([g1, g2, g3], show_png=True)from Assignment_one.performance import EncodingPerformanceAnalysis, combine_analysis_reports
# Analyse the efficient algorithm.
p = EncodingPerformanceAnalysis(efficient=True)
# Store as csv. file.
p.store_metrics()
# Analyse the inefficient algorithm.
p2 = EncodingPerformanceAnalysis(efficient=False)
# Store as csv. file.
p2.store_metrics()
# Load the files and plot their differences.
combine_analysis_reports()