Error in GibbsSampling.sample()

[pmplant]

Subject of the issue

GibbsSampling.sample() is non-functioning, provided example does not work.

Your environment

• pgmpy version
  Name: pgmpy
  Version: 0.1.6

• Python version
  Python 3.6.5

• Operating System
  Linux
  4.15.0-23-generic

Steps to reproduce

Run the example code provided in the definition of GibbsSampling.sample()

from pgmpy.factors.discrete import TabularCPD
from pgmpy.models import BayesianModel
from pgmpy.sampling import GibbsSampling

intel_cpd = TabularCPD('intel', 2, [[0.7], [0.3]])
sat_cpd = TabularCPD('sat', 2, [[0.95, 0.2], [0.05, 0.8]], evidence=['intel'], evidence_card=[2])
student = BayesianModel()
student.add_nodes_from(['intel', 'sat'])
student.add_edge('intel', 'sat')
student.add_cpds(intel_cpd, sat_cpd)
gibbs_chain = GibbsSampling(student)
gibbs_chain.sample(size=3)

Expected behaviour

It should work

Actual behaviour

It don't

Traceback (most recent call last):
  File "gibbexample.py", line 12, in <module>
    gibbs_chain.sample(size=3)
  File "/home/pmp/.local/lib/python3.6/site-packages/pgmpy/sampling/Sampling.py", line 382, in sample
    sampled[0] = np.array([st for var, st in self.state])
ValueError: setting an array element with a sequence.

[ankurankan]

Fixed.