我一直在使用HMMpytk（在Windows操作系统中）开发一个隐马尔可夫模型。在导出转换矩阵时，我得到一个关键错误。我的代码如下：
hmm_model = hmm_faster.HMM()
hmm_model.set_states(['GEN' , 'NICHTGEN'])
hmm_model.set_observations(['x', 'y', 'z'])
hmm_model.randomize_matrices(seed = 10)
start_probability = {'GEN' : startToGene, 'NICHTGEN': startToNotGene}
transition_probability = {'GEN':{'geneToGene':a, 'geneToNotGene':b}, 'NICHTGEN': {'notGeneToGene':c, 'notGeneToNotGene':d}}
hmm_model.set_initial_matrix(start_probability)
hmm_model.set_transition_matrix(transition_probability)
transition_probability中的a，b，c和d是赋值给它们的浮点值的变量。
错误说：
self.trans_matrix[self.st_list_index[st_i]][self.st_list_index[st_j]]=A_matrix[st_i][st_j]
KeyError: 'notGeneToNotGene'