Added multiclass model test.

dbms/tests/external_models/catboost/helpers/server_with_models.py

@@ -20,6 +20,13 @@ CLICKHOUSE_CONFIG = \
     <users_config>users.xml</users_config>
     <tcp_port>{tcp_port}</tcp_port>
     <catboost_dynamic_library_path>{catboost_dynamic_library_path}</catboost_dynamic_library_path>
+    <logger>
+        <level>trace</level>
+        <log>{path}/clickhouse-server.log</log>
+        <errorlog>{path}/clickhouse-server.err.log</errorlog>
+        <size>never</size>
+        <count>50</count>
+    </logger>
 </yandex>
 '''
 

dbms/tests/external_models/catboost/helpers/table.py

@@ -56,7 +56,12 @@ class ClickHouseTable:
         columns = ', '.join(list(float_columns) + list(cat_columns))
         query = "select modelEvaluate('{}', {}) from test.{} format TSV"
         result = self.client.query(query.format(model_name, columns, self.table_name))
-        return tuple(map(float, filter(len, map(str.strip, result.split()))))
+
+        def parse_row(row):
+            values = tuple(map(float, filter(len, map(str.strip, row.replace('(', '').replace(')', '').split(',')))))
+            return values if len(values) != 1 else values[0]
+
+        return tuple(map(parse_row, filter(len, map(str.strip, result.split('\n')))))
 
     def _drop_table(self):
         self.client.query('drop table test.{}'.format(self.table_name))

dbms/tests/external_models/catboost/test_apply_catboost_model/test.py

@@ -234,3 +234,61 @@ def test_apply_float_features_with_mixed_cat_features():
     print 'clickhouse predictions', pred_ch
 
     check_predictions(name, test_target, pred_python, pred_ch, 0.9)
+
+
+def test_apply_multiclass():
+
+    name = 'test_apply_float_features_with_mixed_cat_features'
+
+    train_size = 10000
+    test_size = 10000
+
+    def gen_data(size, seed):
+        data = {
+            'a': generate_uniform_float_column(size, 0., 1., seed + 1),
+            'b': generate_uniform_float_column(size, 0., 1., seed + 2),
+            'c': generate_uniform_string_column(size, ['a', 'b', 'c'], seed + 3),
+            'd': generate_uniform_int_column(size, 1, 4, seed + 4)
+        }
+        return DataFrame.from_dict(data)
+
+    def get_target(df):
+        def target_filter(row):
+            if row['a'] > .3 and row['b'] > .3 and row['c'] != 'a':
+                return 0
+            elif row['a'] * row['b'] > 0.1 and row['c'] != 'b' and row['d'] != 2:
+                return 1
+            else:
+                return 2
+
+        return df.apply(target_filter, axis=1).as_matrix()
+
+    train_df = gen_data(train_size, 42)
+    test_df = gen_data(test_size, 43)
+
+    train_target = get_target(train_df)
+    test_target = get_target(test_df)
+
+    print
+    print 'train target', train_target
+    print 'test target', test_target
+
+    params = {
+        'iterations': 10,
+        'depth': 4,
+        'learning_rate': 1,
+        'loss_function': 'MultiClass'
+    }
+
+    model = train_catboost_model(train_df, train_target, ['c', 'd'], params)
+    pred_python = model.predict(test_df)[:,0].astype(int)
+
+    server = ClickHouseServerWithCatboostModels(name, CLICKHOUSE_TESTS_SERVER_BIN_PATH, PORT)
+    server.add_model(name, model)
+    with server:
+        pred_ch = np.argmax(np.array(server.apply_model(name, test_df, [])), axis=1)
+
+    print 'python predictions', pred_python
+    print 'clickhouse predictions', pred_ch
+
+    check_predictions(name, test_target, pred_python, pred_ch, 0.9)