Merge pull request #661 from jedwin3210/edwin/issue_660

feat(util): add seed parameter for reproducible sampling

Author: pancakereport

CHANGELOG.md

@@ -4,6 +4,11 @@ All notable changes to this project will be documented in this file.
 
 This project adheres to [Semantic Versioning](http://semver.org/).
 
+### v0.18.1
+* Added optional `seed` parameter to `sample_proportions()` for reproducible results
+* Added optional `seed` parameter to `proportions_from_distribution()` for reproducible results
+* Fixed bug where `proportions_from_distribution()` ignored the `column_name` parameter
+
 ### v0.18.0
 * Improved color contrast for charts (gold, blue)
 * Fixed make_array() so it doesn't auto-convert booleans to integers

datascience/util.py

@@ -126,7 +126,7 @@ def plot_normal_cdf(rbound=None, lbound=None, mean=0, sd=1):
 plot_cdf_area = plot_normal_cdf
 
 
-def sample_proportions(sample_size: int, probabilities):
+def sample_proportions(sample_size: int, probabilities, seed=None):
     """Return the proportion of random draws for each outcome in a distribution.
 
     This function is similar to np.random.Generator.multinomial, but returns proportions
@@ -137,15 +137,17 @@ def sample_proportions(sample_size: int, probabilities):
 
         ``probabilities``: An array of probabilities that forms a distribution.
 
+        ``seed``: Optional seed for reproducibility. If None, results will be random.
+
     Returns:
         An array with the same length as ``probability`` that sums to 1.
     """
-    rng = np.random.default_rng()
+    rng = np.random.default_rng(seed)
     return rng.multinomial(sample_size, probabilities) / sample_size
 
 
 def proportions_from_distribution(table, label, sample_size,
-                                  column_name='Random Sample'):
+                                  column_name='Random Sample', seed=None):
     """
     Adds a column named ``column_name`` containing the proportions of a random
     draw using the distribution in ``label``.
@@ -165,6 +167,8 @@ def proportions_from_distribution(table, label, sample_size,
         ``column_name``: The name of the new column that contains the sampled
             proportions. Defaults to ``'Random Sample'``.
 
+        ``seed``: Optional seed for reproducibility. If None, results will be random.
+
     Returns:
         A copy of ``table`` with a column ``column_name`` containing the
         sampled proportions. The proportions will sum to 1.
@@ -173,8 +177,8 @@ def proportions_from_distribution(table, label, sample_size,
         ``ValueError``: If the ``label`` is not in the table, or if
             ``table.column(label)`` does not sum to 1.
     """
-    proportions = sample_proportions(sample_size, table.column(label))
-    return table.with_column('Random Sample', proportions)
+    proportions = sample_proportions(sample_size, table.column(label), seed)
+    return table.with_column(column_name, proportions)
 
 
 def table_apply(table, func, subset=None):

datascience/version.py

@@ -1 +1 @@
-__version__ = '0.18.0'
+__version__ = '0.18.1'

tests/test_util.py

@@ -103,6 +103,30 @@ def test_proportions_from_distribution():
     assert [x in (0, 0.5, 1) for x in ds.sample_proportions(2, ds.make_array(.2, .3, .5))]
 
 
+def test_sample_proportions_seed():
+    """Test seed parameter and backward compatibility"""
+    result1 = ds.sample_proportions(1000, [0.5, 0.5], seed=42)
+    result2 = ds.sample_proportions(1000, [0.5, 0.5], seed=42)
+    assert np.array_equal(result1, result2)
+    
+    result3 = ds.sample_proportions(1000, [0.5, 0.5], seed=99)
+    assert not np.array_equal(result1, result3)
+
+
+def test_proportions_from_distribution_seed_and_column_name():
+    """Test seed parameter and column_name bug fix"""
+    t = ds.Table().with_column('probs', [0.6, 0.4])
+    
+    result1 = ds.proportions_from_distribution(t, 'probs', 1000, seed=42)
+    result2 = ds.proportions_from_distribution(t, 'probs', 1000, seed=42)
+    assert np.array_equal(result1.column(1), result2.column(1))
+    assert _round_eq(1, sum(result1.column(1)))
+    
+    result3 = ds.proportions_from_distribution(t, 'probs', 1000, column_name='My Sample')
+    assert 'My Sample' in result3.labels
+    assert result3.num_columns == 2
+
+
 def test_is_non_string_iterable():
     is_string = 'hello'
     assert ds.is_non_string_iterable(is_string) == False