[Python] How to transform a dict to a tuple without modifying itself [closed]

This is the information I'm using:

customers = [1, 2, 3]
customer_locations = {
1: [[1], [4], [1,4]],
2: [[2], [5], [2,5]],
3: [[3], [6], [3,6]]}


First, I pass the dict history_actual to the function tsp_dinamico then I have to save the cost of that dict in another dic, so I transform the dic to a tuple using function dict_to_tuple_key, but in the next iteration when passing history_actual to the function tsp_dinamico as it expect a dict an error shows up. This is the error:

Cell In[7], line 27
24 # Restricciones
25 # Cada conjunto de clientes debe ser visitado exactamente una vez
26 for c in new_customers:
---> 27 model.addConstr(gp.quicksum(y for i in new_c_locations[c]) == 1, f"Customer_{c}")
29 # Conservación de flujo en los nodos internos
30 for i in data['locations']:

KeyError: (1,)


This is the typical form of new_c_locations (history_actual)

{1: [1, 4], 2: [2], 3: [3]}


it looks like instead of a list now the dict has tuples. I tried with deep copy but it did not work.

for i in range(1, N+1):
with open(f"{Folder}\\instance_base{i}.pkl", 'rb') as f:
data_dummy = pickle.load(f)

V = {}
for t in reversed(customers):
V_t = {}

if t == max(customers):
prev_customers = customers[:-1]
all_histories = list(product(*[customer_locations[c] for c in prev_customers]))
else:
all_histories = V[t+1].keys()

for history in all_histories:
base_history = {c: loc for c, loc in zip(prev_customers, history)}
for location_t in customer_locations[t]:
# Create a deep copy to not modify original dic
history_actual = base_history.copy()
history_actual[t] = location_t


tour, tcost, _, _ = tsp_dinamico(data_dummy, list(history_actual.keys()), history_actual, InitialSolution=True)

# Convert dic to tuple to use as key in V
history_tuple = dict_to_tuple_key(copy.deepcopy(history_actual))
if history_tuple not in V_t or tcost < V_t[history_tuple]:
V_t[history_tuple] = tcost

# Save values in V
V[t] = V_t

best_solution = min(V[1], key=V[1].get)
best_cost = V[1][best_solution]


I need to find a way of passing the dict history_actual to the tsp_dinamico and then transform it to a tuple but keeping it as a dict for the next iteration. I hope this is now a bit clear. Any help will be appreciated.

Continue reading...