Seems 'prob_leave_bandit' in 'state_transition' not used.

[yoongi0428]

For both RecoGym environment (RecoEnv0, RecoEnv1), they initialize 'state_transition' using default arguments passed through config.
It seems that the second row, which correspond to bandit, use 'self.config.prob_leave_organic' instead of 'self.config.prob_leave_bandit'.
The default value of both is the same as 0.1 but I think it might cause unintentional behavior in some cases.
Is this code intended or just a typo?

reco-gym/recogym/envs/reco_env_v1.py

Lines 51 to 58 in f8553d1

	def set_static_params(self):
	# Initialise the state transition matrix which is 3 by 3
	# high level transitions between organic, bandit and leave.
	self.state_transition = np.array([
	[0, self.config.prob_organic_to_bandit, self.config.prob_leave_organic],
	[self.config.prob_bandit_to_organic, 0, self.config.prob_leave_organic],
	[0.0, 0.0, 1.]
	])

[ihtiihti]

You are right, it should be fixed.