Spaces:
				
			
			
	
			
			
		Runtime error
		
	
	
	
			
			
	
	
	
	
		
		
		Runtime error
		
	finished
Browse files- sklearn_proxy.py +16 -1
- tests.py +18 -5
    	
        sklearn_proxy.py
    CHANGED
    
    | @@ -16,6 +16,7 @@ | |
| 16 | 
             
            import evaluate
         | 
| 17 | 
             
            import datasets
         | 
| 18 | 
             
            from sklearn.metrics import get_scorer
         | 
|  | |
| 19 |  | 
| 20 |  | 
| 21 | 
             
            # TODO: Add BibTeX citation
         | 
| @@ -61,6 +62,9 @@ BAD_WORDS_URL = "http://url/to/external/resource/bad_words.txt" | |
| 61 | 
             
            @evaluate.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION)
         | 
| 62 | 
             
            class SklearnProxy(evaluate.Metric):
         | 
| 63 | 
             
                """TODO: Short description of my evaluation module."""
         | 
|  | |
|  | |
|  | |
| 64 |  | 
| 65 | 
             
                def _info(self):
         | 
| 66 | 
             
                    # TODO: Specifies the evaluate.EvaluationModuleInfo object
         | 
| @@ -74,6 +78,7 @@ class SklearnProxy(evaluate.Metric): | |
| 74 | 
             
                        features=datasets.Features({
         | 
| 75 | 
             
                            'predictions': datasets.Value('int64'),
         | 
| 76 | 
             
                            'references': datasets.Value('int64'),
         | 
|  | |
| 77 | 
             
                        }),
         | 
| 78 | 
             
                        # Homepage of the module for documentation
         | 
| 79 | 
             
                        homepage="http://module.homepage",
         | 
| @@ -89,4 +94,14 @@ class SklearnProxy(evaluate.Metric): | |
| 89 |  | 
| 90 | 
             
                def _compute(self, predictions, references, metric_name="accuracy", **kwargs):
         | 
| 91 | 
             
                    scorer = get_scorer(metric_name)
         | 
| 92 | 
            -
             | 
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | 
|  | |
| 16 | 
             
            import evaluate
         | 
| 17 | 
             
            import datasets
         | 
| 18 | 
             
            from sklearn.metrics import get_scorer
         | 
| 19 | 
            +
            from sklearn.base import BaseEstimator
         | 
| 20 |  | 
| 21 |  | 
| 22 | 
             
            # TODO: Add BibTeX citation
         | 
|  | |
| 62 | 
             
            @evaluate.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION)
         | 
| 63 | 
             
            class SklearnProxy(evaluate.Metric):
         | 
| 64 | 
             
                """TODO: Short description of my evaluation module."""
         | 
| 65 | 
            +
                def __init__(self, **kwargs):
         | 
| 66 | 
            +
                    super().__init__(**kwargs)
         | 
| 67 | 
            +
                    self.dummy_estimator = PassThroughtEstimator()
         | 
| 68 |  | 
| 69 | 
             
                def _info(self):
         | 
| 70 | 
             
                    # TODO: Specifies the evaluate.EvaluationModuleInfo object
         | 
|  | |
| 78 | 
             
                        features=datasets.Features({
         | 
| 79 | 
             
                            'predictions': datasets.Value('int64'),
         | 
| 80 | 
             
                            'references': datasets.Value('int64'),
         | 
| 81 | 
            +
             | 
| 82 | 
             
                        }),
         | 
| 83 | 
             
                        # Homepage of the module for documentation
         | 
| 84 | 
             
                        homepage="http://module.homepage",
         | 
|  | |
| 94 |  | 
| 95 | 
             
                def _compute(self, predictions, references, metric_name="accuracy", **kwargs):
         | 
| 96 | 
             
                    scorer = get_scorer(metric_name)
         | 
| 97 | 
            +
             | 
| 98 | 
            +
                    return {metric_name: scorer(self.dummy_estimator, references, predictions, **kwargs)}
         | 
| 99 | 
            +
             | 
| 100 | 
            +
                
         | 
| 101 | 
            +
            class PassThroughtEstimator(BaseEstimator):
         | 
| 102 | 
            +
                def __init__(self):
         | 
| 103 | 
            +
                    pass
         | 
| 104 | 
            +
                def fit(self, X, y):
         | 
| 105 | 
            +
                    return self
         | 
| 106 | 
            +
                def predict(self, X):
         | 
| 107 | 
            +
                    return X
         | 
    	
        tests.py
    CHANGED
    
    | @@ -1,17 +1,30 @@ | |
| 1 | 
            -
             | 
|  | |
|  | |
|  | |
|  | |
| 2 | 
             
                {
         | 
| 3 | 
             
                    "predictions": [0, 0],
         | 
| 4 | 
             
                    "references": [1, 1],
         | 
| 5 | 
            -
                    "result": {" | 
| 6 | 
             
                },
         | 
| 7 | 
             
                {
         | 
| 8 | 
             
                    "predictions": [1, 1],
         | 
| 9 | 
             
                    "references": [1, 1],
         | 
| 10 | 
            -
                    "result": {" | 
| 11 | 
             
                },
         | 
| 12 | 
             
                {
         | 
| 13 | 
             
                    "predictions": [1, 0],
         | 
| 14 | 
             
                    "references": [1, 1],
         | 
| 15 | 
            -
                    "result": {" | 
| 16 | 
             
                }
         | 
| 17 | 
            -
            ]
         | 
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | 
|  | |
| 1 | 
            +
            from sklearn_proxy import SklearnProxy
         | 
| 2 | 
            +
            import unittest
         | 
| 3 | 
            +
             | 
| 4 | 
            +
             | 
| 5 | 
            +
            accuracy_test_cases = [
         | 
| 6 | 
             
                {
         | 
| 7 | 
             
                    "predictions": [0, 0],
         | 
| 8 | 
             
                    "references": [1, 1],
         | 
| 9 | 
            +
                    "result": {"accuracy": 0.0}
         | 
| 10 | 
             
                },
         | 
| 11 | 
             
                {
         | 
| 12 | 
             
                    "predictions": [1, 1],
         | 
| 13 | 
             
                    "references": [1, 1],
         | 
| 14 | 
            +
                    "result": {"accuracy": 1.0}
         | 
| 15 | 
             
                },
         | 
| 16 | 
             
                {
         | 
| 17 | 
             
                    "predictions": [1, 0],
         | 
| 18 | 
             
                    "references": [1, 1],
         | 
| 19 | 
            +
                    "result": {"accuracy": 0.5}
         | 
| 20 | 
             
                }
         | 
| 21 | 
            +
            ]
         | 
| 22 | 
            +
             | 
| 23 | 
            +
             | 
| 24 | 
            +
            class TestGeneral(unittest.TestCase):
         | 
| 25 | 
            +
             | 
| 26 | 
            +
                def test_accuracy(self):
         | 
| 27 | 
            +
                    metric = SklearnProxy()
         | 
| 28 | 
            +
                    for test_case in accuracy_test_cases:
         | 
| 29 | 
            +
                        result = metric.compute(predictions=test_case["predictions"],references=test_case["references"], metric_name="accuracy")
         | 
| 30 | 
            +
                        self.assertEqual(result, test_case["result"])
         | 
