Extract interaction constraint from split evaluator. (#5034)

*  Extract interaction constraints from split evaluator.

The reason for doing so is mostly for model IO, where num_feature and interaction_constraints are copied in split evaluator. Also interaction constraint by itself is a feature selector, acting like column sampler and it's inefficient to bury it deep in the evaluator chain. Lastly removing one another copied parameter is a win.

*  Enable inc for approx tree method.

As now the implementation is spited up from evaluator class, it's also enabled for approx method.

*  Removing obsoleted code in colmaker.

They are never documented nor actually used in real world. Also there isn't a single test for those code blocks.

*  Unifying the types used for row and column.

As the size of input dataset is marching to billion, incorrect use of int is subject to overflow, also singed integer overflow is undefined behaviour. This PR starts the procedure for unifying used index type to unsigned integers. There's optimization that can utilize this undefined behaviour, but after some testings I don't see the optimization is beneficial to XGBoost.

tests/python/test_interaction_constraints.py

@@ -10,7 +10,7 @@ rng = np.random.RandomState(1994)
 
 
 class TestInteractionConstraints(unittest.TestCase):
-    def test_interaction_constraints(self, tree_method='hist'):
+    def run_interaction_constraints(self, tree_method):
         x1 = np.random.normal(loc=1.0, scale=1.0, size=1000)
         x2 = np.random.normal(loc=1.0, scale=1.0, size=1000)
         x3 = np.random.choice([1, 2, 3], size=1000, replace=True)
@@ -25,8 +25,7 @@ class TestInteractionConstraints(unittest.TestCase):
             'eta': 0.1,
             'nthread': 2,
             'interaction_constraints': '[[0, 1]]',
-            'tree_method': tree_method,
-            'verbosity': 2
+            'tree_method': tree_method
         }
         num_boost_round = 12
         # Fit a model that only allows interaction between x1 and x2
@@ -50,8 +49,17 @@ class TestInteractionConstraints(unittest.TestCase):
         diff2 = preds[2] - preds[1]
         assert np.all(np.abs(diff2 - diff2[0]) < 1e-4)
 
+    def test_exact_interaction_constraints(self):
+        self.run_interaction_constraints(tree_method='exact')
+
+    def test_hist_interaction_constraints(self):
+        self.run_interaction_constraints(tree_method='hist')
+
+    def test_approx_interaction_constraints(self):
+        self.run_interaction_constraints(tree_method='approx')
+
     @pytest.mark.skipif(**tm.no_sklearn())
-    def test_training_accuracy(self, tree_method='hist'):
+    def training_accuracy(self, tree_method):
         from sklearn.metrics import accuracy_score
         dtrain = xgboost.DMatrix(dpath + 'agaricus.txt.train?indexing_mode=1')
         dtest = xgboost.DMatrix(dpath + 'agaricus.txt.test?indexing_mode=1')
@@ -73,3 +81,12 @@ class TestInteractionConstraints(unittest.TestCase):
         bst = xgboost.train(params, dtrain, num_boost_round)
         pred_dtest = (bst.predict(dtest) < 0.5)
         assert accuracy_score(dtest.get_label(), pred_dtest) < 0.1
+
+    def test_hist_training_accuracy(self):
+        self.training_accuracy(tree_method='hist')
+
+    def test_exact_training_accuracy(self):
+        self.training_accuracy(tree_method='exact')
+
+    def test_approx_training_accuracy(self):
+        self.training_accuracy(tree_method='approx')