Merge 1dc4842 into efc2a2a

src/cr/cube/measures/pairwise_significance.py

@@ -184,7 +184,7 @@ def t_stats_scale_means(self):
         )
 
         tstats_scale_means = (
-            self._slice.scale_means_row[self._col_idx] - self._slice.scale_means_row
+            self._slice.scale_means_row - self._slice.scale_means_row[self._col_idx]
         ) / (standard_deviation * np.sqrt((1 / counts[self._col_idx]) + (1 / counts)))
 
         return tstats_scale_means

tests/expectations/cat-x-cat-hs-scale-mean-pw-idxs-alt.py

@@ -1,13 +1,13 @@
 (
-    (3, 4, 5, 8, 10),
-    (3, 4, 5, 8, 10),
-    (3, 4, 5, 8, 10),
     (),
     (),
-    (10,),
-    (3, 4, 5, 8, 10),
-    (3, 4, 10),
     (),
-    (4, 10),
+    (0, 1, 2, 6, 7),
+    (0, 1, 2, 6, 7, 9),
+    (0, 1, 2, 6),
     (),
+    (),
+    (0, 1, 2, 6),
+    (),
+    (0, 1, 2, 5, 6, 7, 9),
 )

tests/expectations/cat-x-cat-hs-scale-mean-pw-idxs.py

@@ -1,13 +1,13 @@
 (
-    (3, 4, 5, 8, 10),
-    (3, 4, 10),
-    (3, 4, 5, 8, 10),
     (),
     (),
     (),
-    (3, 4, 5, 8, 10),
-    (3, 4, 10),
+    (0, 1, 2, 6, 7),
+    (0, 1, 2, 6, 7),
+    (0, 2, 6),
     (),
-    (10,),
     (),
+    (0, 2, 6),
+    (),
+    (0, 1, 2, 6, 7, 9),
 )

tests/expectations/cat-x-cat-scale-mean-pw-idxs-alt.py

@@ -1 +1 @@
-((), (0,), (0,))
+((1, 2), (), ())

tests/expectations/cat-x-cat-scale-mean-pw-idxs.py

@@ -1 +1 @@
-((), (0,), (0,))
+((1, 2), (), ())

tests/expectations/cat-x-mr-aug-scale-mean-pw-idxs-alt.py

@@ -1 +1 @@
-((), (0, 4), (0,), (0, 4), ())
+((1, 2, 3), (), (), (), (1, 3))

tests/expectations/cat-x-mr-aug-scale-mean-pw-idxs.py

@@ -1 +1 @@
-((), (0,), (), (0, 4), ())
+((1, 3), (), (), (), (3,))

tests/expectations/cat-x-mr-scale-mean-pw-idxs-alt.py

@@ -1 +1 @@
-((), (0,), (0, 1), (0, 1, 4), (0,))
+((1, 2, 3, 4), (2, 3), (), (), (3,))

tests/expectations/cat-x-mr-scale-mean-pw-idxs.py

@@ -1 +1 @@
-((), (0,), (0, 1), (0, 1, 4), (0,))
+((1, 2, 3, 4), (2, 3), (), (), (3,))

tests/integration/test_pairwise_significance.py

@@ -147,7 +147,7 @@ def test_pairwise_measures_scale_means_nps_type(self):
         assert actual.p_vals_scale_means[0] == 1
         assert actual.t_stats_scale_means[0] == 0
         np.testing.assert_almost_equal(
-            actual.t_stats_scale_means, [0.0, -1.08966143, 1.27412884, 4.14629088]
+            actual.t_stats_scale_means, [0.0, 1.08966143, -1.27412884, -4.14629088]
         )
         np.testing.assert_almost_equal(
             actual.p_vals_scale_means,
@@ -164,7 +164,7 @@ def test_pairwise_measures_scale_means_nps_type(self):
             [2.7638510e-01, 1.0000000e00, 1.4344585e-03, 1.7905266e-09],
         )
         np.testing.assert_almost_equal(
-            actual.t_stats_scale_means, [1.08966143, 0.0, 3.19741668, 6.10466696]
+            actual.t_stats_scale_means, [-1.08966143, 0.0, -3.19741668, -6.10466696]
         )
 
     def test_ttests_scale_means_cat_x_cat_pruning_and_hs(self):
@@ -177,7 +177,7 @@ def test_ttests_scale_means_cat_x_cat_pruning_and_hs(self):
 
         np.testing.assert_almost_equal(
             actual.t_stats_scale_means,
-            [0.0, -1.64461503, -1.92387847, np.nan, -1.06912069],
+            [0.0, 1.64461503, 1.92387847, np.nan, 1.06912069],
         )
         np.testing.assert_almost_equal(
             actual.p_vals_scale_means, [1.0, 0.1046981, 0.059721, np.nan, 0.2918845]
@@ -189,7 +189,7 @@ def test_ttests_scale_means_cat_x_cat_pruning_and_hs(self):
 
         np.testing.assert_almost_equal(
             actual.t_stats_scale_means,
-            [0.0, -0.9387958, -1.644615, -1.9238785, np.nan, -1.0691207],
+            [0.0, 0.9387958, 1.644615, 1.9238785, np.nan, 1.0691207],
         )
 
         np.testing.assert_almost_equal(
@@ -207,7 +207,7 @@ def test_ttests_scale_means_cat_x_cat_pruning_and_hs(self):
 
         np.testing.assert_almost_equal(
             actual.t_stats_scale_means,
-            [0.0, -0.93879579, -1.64461503, -1.92387847, -1.06912069],
+            [0.0, 0.93879579, 1.64461503, 1.92387847, 1.06912069],
         )
         np.testing.assert_almost_equal(
             actual.p_vals_scale_means, [1.0, 0.3500141, 0.1046981, 0.059721, 0.2918845]
@@ -222,7 +222,7 @@ def test_ttests_scale_means_cat_x_cat_pruning_and_hs(self):
         actual = slice_.pairwise_significance_tests[0]
 
         np.testing.assert_almost_equal(
-            actual.t_stats_scale_means, [0.0, -1.64461503, -1.92387847, -1.06912069]
+            actual.t_stats_scale_means, [0.0, 1.64461503, 1.92387847, 1.06912069]
         )
         np.testing.assert_almost_equal(
             actual.p_vals_scale_means, [1.0, 0.1046981, 0.059721, 0.2918845]
@@ -236,16 +236,16 @@ def test_pairwise_t_stats_scale_means_with_hs(self):
             actual.t_stats_scale_means,
             [
                 0.0,
-                0.70543134,
-                0.06749456,
-                3.59627589,
-                4.15672996,
-                3.26353764,
-                0.30034779,
-                0.8383222,
-                3.02379687,
-                0.85672124,
-                4.26974185,
+                -0.70543134,
+                -0.06749456,
+                -3.59627589,
+                -4.15672996,
+                -3.26353764,
+                -0.30034779,
+                -0.8383222,
+                -3.02379687,
+                -0.85672124,
+                -4.26974185,
             ],
         )
 
@@ -271,7 +271,7 @@ def test_ttests_scale_means_use_unweighted_n_for_variance(self):
         actual = slice_.pairwise_significance_tests[0]
 
         np.testing.assert_almost_equal(
-            actual.t_stats_scale_means, [0.0, 4.38871748, 3.99008596, 5.15679647]
+            actual.t_stats_scale_means, [0.0, -4.38871748, -3.99008596, -5.15679647]
         )
         np.testing.assert_almost_equal(
             actual.p_vals_scale_means,