fix (#706)

Author: fealho

rdt/transformers/numerical.py

@@ -252,9 +252,10 @@ class GaussianNormalizer(FloatFormatter):
                 * ``gamma``: Use a Gamma distribution.
                 * ``beta``: Use a Beta distribution.
                 * ``student_t``: Use a Student T distribution.
-                * ``gussian_kde``: Use a GaussianKDE distribution. This model is non-parametric,
+                * ``gaussian_kde``: Use a GaussianKDE distribution. This model is non-parametric,
                   so using this will make ``get_parameters`` unusable.
                 * ``truncated_gaussian``: Use a Truncated Gaussian distribution.
+                # ``uniform``: Use a UniformUnivariate distribution.
 
         missing_value_generation (str or None):
             The way missing values are being handled. There are three strategies:
@@ -269,24 +270,6 @@ class GaussianNormalizer(FloatFormatter):
 
     _univariate = None
 
-    def __init__(self, model_missing_values=None, learn_rounding_scheme=False,
-                 enforce_min_max_values=False, distribution='truncated_gaussian',
-                 missing_value_generation='random'):
-        super().__init__(
-            model_missing_values=model_missing_values,
-            missing_value_generation=missing_value_generation,
-            learn_rounding_scheme=learn_rounding_scheme,
-            enforce_min_max_values=enforce_min_max_values
-        )
-
-        self.distribution = distribution  # Distribution initialized by the user
-
-        self._distributions = self._get_distributions()
-        if isinstance(distribution, str):
-            distribution = self._distributions[distribution]
-
-        self._distribution = distribution
-
     @staticmethod
     def _get_distributions():
         try:
@@ -305,8 +288,25 @@ def _get_distributions():
             'student_t': univariate.StudentTUnivariate,
             'gaussian_kde': univariate.GaussianKDE,
             'truncated_gaussian': univariate.TruncatedGaussian,
+            'uniform': univariate.UniformUnivariate,
         }
 
+    def __init__(self, model_missing_values=None, learn_rounding_scheme=False,
+                 enforce_min_max_values=False, distribution='truncated_gaussian',
+                 missing_value_generation='random'):
+        super().__init__(
+            model_missing_values=model_missing_values,
+            missing_value_generation=missing_value_generation,
+            learn_rounding_scheme=learn_rounding_scheme,
+            enforce_min_max_values=enforce_min_max_values
+        )
+
+        self._distributions = self._get_distributions()
+        if isinstance(distribution, str):
+            distribution = self._distributions[distribution]
+
+        self._distribution = distribution
+
     def _get_univariate(self):
         distribution = self._distribution
         if any(isinstance(distribution, dist) for dist in self._distributions.values()):

tests/integration/transformers/test_numerical.py

@@ -1,5 +1,6 @@
 import numpy as np
 import pandas as pd
+from copulas import univariate
 
 from rdt.transformers.numerical import ClusterBasedNormalizer, FloatFormatter, GaussianNormalizer
 
@@ -195,6 +196,66 @@ def test_int_nan(self):
         reverse = ct.reverse_transform(transformed)
         np.testing.assert_array_almost_equal(reverse, data, decimal=2)
 
+    def test_uniform(self):
+        """Test it works when distribution='uniform'."""
+        # Setup
+        data = pd.DataFrame(np.random.uniform(size=1000), columns=['a'])
+        ct = GaussianNormalizer(distribution='uniform')
+
+        # Run
+        ct.fit(data, 'a')
+        transformed = ct.transform(data)
+        reverse = ct.reverse_transform(transformed)
+
+        # Assert
+        assert isinstance(transformed, pd.DataFrame)
+        assert transformed.shape == (1000, 1)
+
+        np.testing.assert_almost_equal(transformed['a'].mean(), 0, decimal=1)
+        np.testing.assert_almost_equal(transformed['a'].std(), 1, decimal=1)
+
+        np.testing.assert_array_almost_equal(reverse, data, decimal=1)
+
+    def test_uniform_object(self):
+        """Test it works when distribution=UniformUnivariate()."""
+        # Setup
+        data = pd.DataFrame(np.random.uniform(size=1000), columns=['a'])
+        ct = GaussianNormalizer(distribution=univariate.UniformUnivariate())
+
+        # Run
+        ct.fit(data, 'a')
+        transformed = ct.transform(data)
+        reverse = ct.reverse_transform(transformed)
+
+        # Assert
+        assert isinstance(transformed, pd.DataFrame)
+        assert transformed.shape == (1000, 1)
+
+        np.testing.assert_almost_equal(transformed['a'].mean(), 0, decimal=1)
+        np.testing.assert_almost_equal(transformed['a'].std(), 1, decimal=1)
+
+        np.testing.assert_array_almost_equal(reverse, data, decimal=1)
+
+    def test_uniform_class(self):
+        """Test it works when distribution=UniformUnivariate."""
+        # Setup
+        data = pd.DataFrame(np.random.uniform(size=1000), columns=['a'])
+        ct = GaussianNormalizer(distribution=univariate.UniformUnivariate)
+
+        # Run
+        ct.fit(data, 'a')
+        transformed = ct.transform(data)
+        reverse = ct.reverse_transform(transformed)
+
+        # Assert
+        assert isinstance(transformed, pd.DataFrame)
+        assert transformed.shape == (1000, 1)
+
+        np.testing.assert_almost_equal(transformed['a'].mean(), 0, decimal=1)
+        np.testing.assert_almost_equal(transformed['a'].std(), 1, decimal=1)
+
+        np.testing.assert_array_almost_equal(reverse, data, decimal=1)
+
 
 class TestClusterBasedNormalizer:
 

tests/unit/transformers/test_numerical.py

@@ -841,6 +841,7 @@ def test__get_distributions(self):
             'student_t': univariate.StudentTUnivariate,
             'gaussian_kde': univariate.GaussianKDE,
             'truncated_gaussian': univariate.TruncatedGaussian,
+            'uniform': univariate.UniformUnivariate
         }
         assert distributions == expected