Minimum degree of a graph, Python

Question

Why the error when in 'keys'?? I'm trying to calculate the minimum degree of a graph Operations in undirected graph Let me know what's wrong pls

graph = dict()

def add_node(v):
    if v not in graph:
        graph[v] = dict()

def add_edge(v1,v2,cost):
    if v1 in grafo and v2 in graph:    
        graph[v1][v2] = cost
        graph[v2][v1] = cost
        
def del_edge(v1,v2):
    if v1 in graph and v2 in graph[v1]:
        graph[v1].pop(v2)
        graph[v2].pop(v1)

def del_node(v):
    if v not in graph:
        return       
    for v2 in graph[v]:
        graph[v2].pop(v)
    graph.pop(v)
    
def nodes(self):
    return iter(self._graph)

def neighbour_count(self, n):
    return len(self.graph[n])



    
n = int(input())
for _ in range(n):
    comands = input()
    oper,v1,v2,peso,*_ = comands.split(" ") + ["",""]
    if   oper == 'IV': add_node(v1)
    elif oper == 'IA': add_edge(v1,v2,float(peso))
    elif oper == 'RV': del_node(v1)
    elif oper == 'RA': del_edge(v1,v2)

keys = [i for i in graph.nodes()]
degrees = [neighbour_count(i) for i in keys]

print(min(degrees))

Example input:

7
IV A
IV B
IA A B
IV C
IA A C
IA C B
RA A B

output:

1

os vértices A e B só tem uma aresta enquanto C tem duas

Answer 1

graph is a dictionary, and dictionaries don't have a nodes method. So the actual mistake would be in the definition of keys,

[i for i in graph.nodes()] .

You've done def nodes(self): return iter(self._graph) as if you're writing methods for a custom class, but there are no custom classes in your code, you're just using an instance of the native dict class. Which again, doesn't have a nodes method.

You have a couple of good options here:

1. Define a custom class that contains an instance of the native dictionary class. This way you can use its functionality without inheriting its complexity
2. Forget about OOP and implement nodes (and whatever else) as a function instead of a method

Other options include inheriting from dictionary or implementing a hash table from scratch, neither of which are ideal.