print(f"Value of the configuration after Hill Climber: "
f"{climber.graph.calculate_value()}")
# --------------------------- Simulated Annealing --------------------------
# It is very difficult to find a good starting temperature for SA. A rule to
# help you find one is to use the maximum change in score that could happen
# when mutating your state. In our case, this is 19, because the transmitter
# maximum difference in score between the most expensive and the cheapest
# transmitter is 19.
print("Setting up Simulated Annealing...")
simanneal = sa.SimulatedAnnealing(random_graph,
transmitters.get_scheme(1),
temperature=19)
print("Running Simulated Annealing...")
simanneal.run(2000, verbose=True)
print(f"Value of the configuration after Simulated Annealing: "
f"{simanneal.graph.calculate_value()}")
# --------------------------- Visualisation --------------------------------
# Turn fast_plot on for a matplotlib plot, which will be faster than Bokeh.
# Bokeh will also be too slow for plotting Russia, so turn on when Russia is
# selected as input!
fast_plot = False
vis.visualise(simanneal.graph,
f"data/{data_folder}/{data_folder}_regions.shp",
fast_plot=fast_plot)
from code.algorithms import randomize
from code.classes import graph, transmitters
from code.visualisation import visualise as vis
if __name__ == '__main__':
# Create a graph from our data
test_graph = graph.Graph('data/US/states.csv', 'data/US/neighbours.csv')
# Create the transmitter cost schemes
transmitters = transmitters.CostScheme('data/transmitters.csv')
# Perform algorithm that keeps reassigning values until the case is solvedd
randomize.random_reassignment(test_graph, transmitters.get_scheme(1))
# Example on how to get the nodes that have neighbours with the same value and print them
violating_nodes = test_graph.get_violations()
print(len(violating_nodes))
print(test_graph.get_violations())
vis.visualise(test_graph, "data/US/gz_2010_us_040_00_500k.json")
climber = hc.HillClimber(random_graph, transmitters.get_scheme(1))
print("Running Hill Climber...")
climber.run(2000, verbose=True)
print(
f"Value of the configuration after Hill Climber: {climber.graph.calculate_value()}"
)
# --------------------------- Simulated Annealing --------------------------
# It is very difficult to find a good starting temperature for SA. A rule to
# help you find one is to use the maximum change in score that could happen
# when mutating your state. In our case, this is 19, because the transmitter
# maximum difference in score between the most expensive and the cheapest
# transmitter is 19.
print("Setting up Simulated Annealing...")
simanneal = sa.SimulatedAnnealing(random_graph,
transmitters.get_scheme(1),
temperature=19)
print("Running Simulated Annealing...")
simanneal.run(2000, verbose=True)
print(
f"Value of the configuration after Simulated Annealing: {simanneal.graph.calculate_value()}"
)
# --------------------------- Visualisation --------------------------------
vis.visualise(simanneal.graph, "data/US/gz_2010_us_040_00_500k.json")
swap_answers = prompt(swap_type, style=style)
# With Swap
if swap_answers["swap"]:
greedy_swap = greedy.SwapGreedy(district, 1, 0)
result = greedy_swap.run_battery_swap()
else:
greedy = greedy.Greedy(district, 1, 0)
result = greedy.run_battery()
# Greedy House
else:
greedy_swap = greedy.SwapGreedy(district, 1, 0)
result = greedy_swap.run_houses_swap()
# Clustering algorithms
else:
clust = cluster.Cluster(district)
result = clust.run_cluster()
# Table overview of results
x.add_column("District", [result["district"].name])
x.add_column("Success", [result["success"]])
x.add_column("Total Cost", [result["district"].total_cost])
x.add_column("Discounted Cost", [result["district"].discounted_cost])
print(x)
# Visualisation
vis.visualise(result["district"])