Home | History | Annotate | Download | only in kernel_tests 
      1 # Copyright 2015 The TensorFlow Authors. All Rights Reserved.
      2 #
      3 # Licensed under the Apache License, Version 2.0 (the "License");
      4 # you may not use this file except in compliance with the License.
      5 # You may obtain a copy of the License at
      6 #
      7 #     http://www.apache.org/licenses/LICENSE-2.0
      8 #
      9 # Unless required by applicable law or agreed to in writing, software
     10 # distributed under the License is distributed on an "AS IS" BASIS,
     11 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
     12 # See the License for the specific language governing permissions and
     13 # limitations under the License.
     14 # ==============================================================================
     15 """Tests for PrecisionOp."""
     16 
     17 from __future__ import absolute_import
     18 from __future__ import division
     19 from __future__ import print_function
     20 
     21 import numpy as np
     22 
     23 from tensorflow.python.framework import constant_op
     24 from tensorflow.python.framework import errors_impl
     25 from tensorflow.python.ops import nn_ops
     26 from tensorflow.python.platform import test
     27 
     28 
     29 class InTopKTest(test.TestCase):
     30 
     31   def _validateInTopK(self, predictions, target, k, expected):
     32     np_ans = np.array(expected)
     33     with self.test_session():
     34       precision = nn_ops.in_top_k(predictions, target, k)
     35       out = precision.eval()
     36       self.assertAllClose(np_ans, out)
     37       self.assertShapeEqual(np_ans, precision)
     38 
     39   def testInTop1(self):
     40     predictions = [[0.1, 0.3, 0.2, 0.4], [0.1, 0.2, 0.3, 0.4]]
     41     target = [3, 1]
     42     self._validateInTopK(predictions, target, 1, [True, False])
     43 
     44   def testInTop2(self):
     45     predictions = [[0.1, 0.3, 0.2, 0.4], [0.1, 0.2, 0.3, 0.4]]
     46     target = [0, 2]
     47     self._validateInTopK(predictions, target, 2, [False, True])
     48 
     49   def testInTop2Tie(self):
     50     # Class 2 and 3 tie for 2nd, so both are considered in top 2.
     51     predictions = [[0.1, 0.3, 0.2, 0.2], [0.1, 0.3, 0.2, 0.2]]
     52     target = [2, 3]
     53     self._validateInTopK(predictions, target, 2, [True, True])
     54 
     55   def testInTop2_int64Target(self):
     56     predictions = [[0.1, 0.3, 0.2, 0.4], [0.1, 0.2, 0.3, 0.4]]
     57     target = np.asarray([0, 2]).astype(np.int64)
     58     self._validateInTopK(predictions, target, 2, [False, True])
     59 
     60   def testInTopNan(self):
     61     predictions = [[0.1, float("nan"), 0.2, 0.4], [0.1, 0.2, 0.3, float("inf")]]
     62     target = [0, 2]
     63     self._validateInTopK(predictions, target, 2, [False, False])
     64 
     65   def testBadTarget(self):
     66     predictions = [[0.1, 0.3, 0.2, 0.4], [0.1, 0.2, 0.3, 0.4]]
     67     target = [0, 80000]
     68     with self.test_session():
     69       with self.assertRaisesRegexp(errors_impl.InvalidArgumentError,
     70                                    "target.*out of range"):
     71         nn_ops.in_top_k(predictions, target, 2).eval()
     72 
     73   def testTensorK(self):
     74     predictions = [[0.1, 0.3, 0.2, 0.4], [0.1, 0.2, 0.3, 0.4]]
     75     target = [0, 2]
     76     k = constant_op.constant(3)
     77     np_ans = np.array([False, True])
     78     with self.test_session():
     79       precision = nn_ops.in_top_k(predictions, target, k)
     80       out = precision.eval()
     81       self.assertAllClose(np_ans, out)
     82       self.assertShapeEqual(np_ans, precision)
     83 
     84 if __name__ == "__main__":
     85   test.main()
     86