attila.py

# att_effect_economy_gdp_industry
# att_bld_roman_west_city_major_1
import pathlib

from engine.enums import NameType, ProblemState, SolverType
from engine.games import Games
from engine.models.game_attila import AttilaGame
from engine.problem import Problem

# PuLP is a linear and mixed integer programming modeler written in Python.

"""
A province contains n regions.
A capital and smaller settlements are regions part of a province.
A region contains buildings.

A province has public order, global economy, religion, and food, which is the total of all regions.
A region has sanitation, public order, economy, religion, and food.

We need to optimize using linear programming the economy of a province, here using PuLP.
"""

"""
('cha_effect_imperium', 'faction_to_faction_own_unseen', '10.0000'), ('att_effect_economy_gdp_trade_local', 'building_to_building_own', '200.0000'), ('att_effect_public_order_base', 'province_to_province_own', '2.0000'), ('att_effect_region_resource_wine_production', 'building_to_building_own', '10.0000'), ('att_effect_economy_gdp_trade_local', 'building_to_building_own', '300.0000'), ('att_effect_public_order_base', 'province_to_province_own', '4.0000'), ('att_effect_region_resource_wine_production', 'building_to_building_own', '15.0000'), ('att_effect_economy_gdp_trade_local', 'building_to_building_own', '400.0000'), ('att_effect_public_order_base', 'province_to_province_own', '6.0000'), ('att_effect_region_resource_wine_production', 'building_to_building_own', '20.0000'), ('att_effect_economy_gdp_trade_local', 'building_to_building_own', '200.0000'), ('att_effect_province_growth_building', 'province_to_province_own', '2.0000'), ('att_effect_economy_gdp_trade_local', 'building_to_building_own', '300.0000'), ('att_effect_province_growth_building', 'province_to_province_own', '4.0000'), ('att_effect_economy_gdp_trade_local', 'building_to_building_own', '400.0000'), ('att_effect_province_growth_building', 'province_to_province_own', '6.0000'), ('att_effect_economy_gdp_mod_industry', 'province_to_region_own', '10.0000'), ('att_effect_region_resource_wood_production', 'building_to_building_own', '10.0000'), ('att_effect_economy_gdp_mod_industry', 'province_to_region_own', '20.0000'), ('att_effect_region_resource_wood_production', 'building_to_building_own', '15.0000'), ('att_effect_economy_gdp_mod_industry', 'province_to_region_own', '30.0000'), ('att_effect_region_resource_wood_production', 'building_to_building_own', '20.0000')]})
"""

if __name__ == '__main__':
    # Create a province, with regions, and buildings constraints.
    Games.instance = AttilaGame(campaign=AttilaGame.Campaign.ATTILA,
                                faction=AttilaGame.Factions.ATT_FACT_EASTERN_ROMAN_EMPIRE,
                                religion=AttilaGame.Religion.CHRIST_ORTHODOX)
    parser = Games.instance.get_parser()
    Games.buildings = parser.buildings
    # Games.instance.campaign = AttilaGame.Campaign.CHARLEMAGNE

    # Attila data folder
    path = pathlib.Path(__file__).parent.absolute() / "data" / "attila"
    parser.parse_building_effects_junction_tables()
    parser.parse_start_pos_tsv(path)

    # Linear programming problem
    lp_problem = Problem(solver=SolverType.GOOGLE)
    # lp_problem.add_provinces()

    # Options
    Games.USE_NAME = NameType.NAME

    for province in parser.provinces.values():
        lp_problem.reset_problem()
        lp_problem.add_province(province)
        # Filter out all buildings
        for region in province.regions:
            region.add_buildings()
            # Filter out city build below x (to force a city level)
            region.filter_city_level(4)
            region.filter_building_level(4)
            region.filter_military()
            # Necessary (for now) for Google OR-Tools to add the variables which were added lazily
            lp_problem.problem.filter_added({building.name: building for building in region.buildings})
        lp_problem.state = ProblemState.FILTERS_ADDED

        # Set province wide fertility : impacts food and GDP
        Games.fertility = 0

        # Regional constraints
        for region in province.regions:
            region.add_constraints()
        # Sanitation is regional, but requires province wide view to look at province wide effects
        province.add_sanitation_constraint()

        # Province constraints
        province.add_food_constraint()
        province.add_public_order_constraint()
        lp_problem.state = ProblemState.CONSTRAINTS_ADDED

        # GDP maximization
        lp_problem.add_objective()
        # lp_problem.print_problem_xy()
        lp_problem.solve()
        print(f"{province.get_name_output()} : {lp_problem.problem.get_objective()}")

        # Print the variables equal to 1 with their respective contribution
        # if province.print_name == "Thracia":
        #     lp_problem.print_problem_answers()
        # lp_problem.print_problem_answers()
        province.clean()

    print(f"Total solving time: {lp_problem.global_time / 1_000_000_000} seconds")