If hypothesis="1 2 1 2" reference="1 2 3 4 5 1 2 6", what is the right result?

[DtYXs]

from rouge import Rouge

hypothesis="1 2 1 2"
reference="1 2 3 4 5 1 2 6"
rouge = Rouge()
scores = rouge.get_scores(hypothesis, reference)

so, what is the right result of scores?

When I use pip install rouge to install, the result:

[{'rouge-1': {'f': 0.6666666622222223, 'p': 1.0, 'r': 0.5},
  'rouge-2': {'f': 0.3999999958, 'p': 0.6666666666666666, 'r': 0.2857142857142857},
  'rouge-l': {'f': 0.4999999962500001, 'p': 1.0, 'r': 0.3333333333333333}}]

When I use git clone xxx python setup.py install to install, the result:

[{'rouge-1': {'f': 0.4999999962500001, 'p': 1.0, 'r': 0.3333333333333333},
  'rouge-2': {'f': 0.24999999625000005, 'p': 0.5, 'r': 0.16666666666666666},
  'rouge-l': {'f': 0.4999999962500001, 'p': 1.0, 'r': 0.3333333333333333}}]

I want to know which is the right result?

Thanks

[sanghyuk-choi]

@DtYXs // The Official ROUGE Script (ROUGE1.5.5.pl) results

1 ROUGE-1 Average_R: 0.50000 (95%-conf.int. 0.50000 - 0.50000)
1 ROUGE-1 Average_P: 1.00000 (95%-conf.int. 1.00000 - 1.00000)
1 ROUGE-1 Average_F: 0.66667 (95%-conf.int. 0.66667 - 0.66667)
---------------------------------------------
1 ROUGE-2 Average_R: 0.28571 (95%-conf.int. 0.28571 - 0.28571)
1 ROUGE-2 Average_P: 0.66667 (95%-conf.int. 0.66667 - 0.66667)
1 ROUGE-2 Average_F: 0.40000 (95%-conf.int. 0.40000 - 0.40000)
---------------------------------------------
1 ROUGE-L Average_R: 0.50000 (95%-conf.int. 0.50000 - 0.50000)
1 ROUGE-L Average_P: 1.00000 (95%-conf.int. 1.00000 - 1.00000)
1 ROUGE-L Average_F: 0.66667 (95%-conf.int. 0.66667 - 0.66667)

So. I think pip version is the right result.