setting binary initial_state results in wrong solution

[j-i-k-o]

Describe the bug
setting initial_state keyword of sample_qubo with binary variables (0, 1) results in weird answer.

To Reproduce

>>> import openjij as oj
>>> a = oj.SASampler()
>>> a.sample_qubo(Q={(2, 2): -68.55, (0, 2): 0.1, (1, 3): 0.1, (3, 4): 0.1, (1, 4): 0.1, (4, 4): -66.55, (2, 3): 0.1, (1, 1): -69.55, (0, 0): -70.55, (0, 4): 0.1, (0, 3): 0.1, (2, 4): 0.1, (0, 1): 0.1, (3, 3): -67.55, (1, 2): 0.1}, num_sweeps=1000, initial_state=[1,1,0,1,1])
Response(rec.array([([1, 1, 0, 1, 1], -273.6, 1)],
          dtype=[('sample', 'i1', (5,)), ('energy', '<f8'), ('num_occurrences', '<i8')]), Variables([0, 1, 2, 3, 4]), {'system': [], 'sampling_time': 1059.5160019875038, 'execution_time': 900.4790008475538, 'list_exec_times': array([900.47900085]), 'schedule': {'beta_max': 184.20680743952366, 'beta_min': 0.01964981376498782, 'num_sweeps': 1000}}, 'BINARY')

Expected behavior
Result binary array should be [1,1,1,1,1]

Additional context
This error comes from the missing conversion of the binary initial_state array to spin array.

[j-i-k-o]

solved in the latest version