Traveller salesman backtracking

[synced]

So I have the traveller salesman (backtracking) with a cost on the edges (I consider a full graph with 0 on the main diagonal and a cost in the rest, visiting every node just one time except starting node ), the problem is that my algorithm calculates the cost well ( cost = 13 ) but print 0 - 3 - 3 - 3 - 0 instead of the real solution ( I considered a 4 x 4 matrix starting from node 0 ) I am beginner in python, any help?

    def posible(k):
        global c
        if  k != n:
            for i in range(0,k-1):
                if x[k] == x[i]:
                    return False
        else:
            c = a[x[n-1]][x[n]]
            if x[k] != x[0]:
            return False

            for i in range(1,k):
                if x[k] == x[i]:
                    return False
                c = c + a[x[i-1]][x[i]]

        if k > 0:
            if a[x[k-1]][x[k]] == 0:
                return False

        return True

    def compare():
        global cmin
        global o
        global c
        if c < cmin:
            cmin = c
            o = x

    def back(k):
        for i in range(0,n):
            x[k] = i
            if posible(k) == True:
                if k == n:
                    compare()
                else:
                    back(k+1)


    n = 4
    a = [ [0,1,2,3],
          [1,0,4,5],
          [2,4,0,5],
          [3,5,5,0]
        ]
    c = 100
    x = [ 0, 0 , 0 , 0 , 0 ]
    o = [ 0, 0 , 0 , 0 , 0 ]
    cmin = 1000
    x[1] = 0

    back(1)
    o.append(0)
    if cmin == 1000:
        print('No solution')
    else:
        for i in range(0,n):
            print(o[i],end= ' ')
        print('0', end=' ')
        print(" cost : " + str(cmin))