def input_greedy(batteries, houses, command):
"""
This algorithm chooses randomly a house and connects it to a battery based
on best (shortest) distance. It asks for number of repeats to run the
algorithm. It saves all solutions to one .csv file and it writes the
best price of all repeats to another .csv file. The best price is
also visualized.
"""
dist, distdict, lowbprice = distancearr(batteries, houses)
print("run how many times?")
repeats = int(input("> "))
best = Node()
best.price = 100000
for i in range(0, repeats):
greedy(dist, batteries, houses)
price = price_calc(batteries, distdict)
if price < best.price:
best.batts = [[], [], [], [], []]
best.fillnode(batteries, houses, price)
write_to_csv(argv[1], command, price)
reset(batteries, houses)
savefig = bestplot(argv[1], command, best.price)
print(f"Best price found: {best.price}")
if savefig is True:
nodetoclasses(batteries, houses, best)
visualize(batteries, houses, argv[1], command, best.price)
def input_kmeans(batteries, houses, command):
"""
this algorithm takes the desired amount of clusters between 5 and 17 as
input. The algorithm returns a grid with the desired amount of clusters
on (local) optimal positions.
"""
bestprice = 100000
for k in range(5, 17):
clusters, connectedhomes = KmeansClusterdistance(houses, batteries, k)
b, h = clustertoclasses(batteries, houses, clusters, connectedhomes)
if b is False:
k -= 1
continue
batteries = b
houses = h
dist, distdict, lowbprice = distancearr(batteries, houses)
price = price_calc(batteries, distdict)
if price < bestprice:
print(price)
print(k)
bestprice = price
bestbat = batteries
batteries = bestbat
visualize(batteries, houses, argv[1], command, bestprice)
def input_hillclimber(batteries, houses, command):
"""
Iterates over all houses and checks if an improvement can be made by
switching the connections of 2 houses. It asks for number of repeats to run
the algorithm. It saves all solutions to one .csv file and it writes the
best price of all repeats to another .csv file. The best price is
also visualized.
"""
dist, distdict, lowbprice = distancearr(batteries, houses)
print("base: greedy or random")
base = (input("> ")).lower()
if base != "greedy" and base != "random":
return False
command = base + "_" + command
print("run how many times?")
repeats = int(input("> "))
best = Node()
best.price = 100000
for i in range(0, repeats):
if base == "greedy":
greedy(dist, batteries, houses)
elif base == "random":
random_alg(distdict, batteries, houses)
hillclimber(dist, distdict, batteries, houses)
price = price_calc(batteries, distdict)
if price < best.price:
best.batts = [[], [], [], [], []]
best.fillnode(batteries, houses, price)
write_to_csv(argv[1], command, price)
reset(batteries, houses)
savefig = bestplot(argv[1], command, best.price)
print(f"Best price found: {best.price}")
nodetoclasses(batteries, houses, best)
if savefig is True:
visualize(batteries, houses, argv[1], command, best.price)
battery_optimization(batteries)
dist, distdict, lowbprice = distancearr(batteries, houses)
price = price_calc(batteries, distdict)
print(f"Best price found with optimize: {price}")
visualize(batteries, houses, argv[1], command, price)
def input_bfs(batteries, houses, command):
"""
Only neighbourhood 4 should be tried, because the state space of the other
neighbourhoods is too big.
"""
node = Node()
best = Node()
dist, distdict, lowbprice = distancearr(batteries, houses)
best.price = 100000
best = bfs(node, batteries, houses, distdict, best)
nodetoclasses(batteries, houses, best)
price = price_calc(batteries, distdict)
print(f"Best price found: {price}")
visualize(batteries, houses, argv[1], command)
def input_randclimber(batteries, houses, command):
"""
A version of hillclimber which switches 2 random houses until no
improvements are made for the set number of times. It asks for number of
repeats to run the algorithm. It saves all solutions to one .csv file and
it writes the best price of all repeats to another .csv file. The best
price is also visualized.
"""
dist, distdict, lowbprice = distancearr(batteries, houses)
print("base: greedy or random")
base = (input("> ")).lower()
if base != "greedy" and base != "random":
return False
print("repeat until no change for how many times?")
repetitions = int(input("> "))
command = base + "_" + command + str(repetitions)
print("run how many times?")
repeats = int(input("> "))
best = Node()
best.price = 100000
for i in range(0, repeats):
if base == "greedy":
greedy(dist, batteries, houses)
elif base == "random":
random_alg(distdict, batteries, houses)
randclimber(repetitions, distdict, batteries, houses)
price = price_calc(batteries, distdict)
if price < best.price:
best.batts = [[], [], [], [], []]
best.fillnode(batteries, houses, price)
write_to_csv(argv[1], command, price)
reset(batteries, houses)
savefig = bestplot(argv[1], command, best.price)
print(best.price)
if savefig is True:
nodetoclasses(batteries, houses, best)
visualize(batteries, houses, argv[1], command, best.price)