def test_benchmark_result_initialization():
br = BenchmarkResult(["loop1", "loop2"], 5, ["mae", "mse"])
for loop in ["loop1", "loop2"]:
for metric in ["mae", "mse"]:
metric_data = br.extract_metric_as_array(loop, metric)
assert metric_data.size == 0
assert metric_data.shape == (5, 0)
def test_benchmark_result_initialization():
br = BenchmarkResult(['loop1', 'loop2'], 5, ['mae', 'mse'])
for loop in ['loop1', 'loop2']:
for metric in ['mae', 'mse']:
metric_data = br.extract_metric_as_array(loop, metric)
assert metric_data.size == 0
assert metric_data.shape == (5, 0)
def test_benchmark_result_extract_invalid_names():
br = BenchmarkResult(['loop1', 'loop2'], 5, ['mae', 'mse'])
with pytest.raises(KeyError):
br.extract_metric_as_array('invalid loop', 'mse')
with pytest.raises(KeyError):
br.extract_metric_as_array('loop1', 'invalid metric')
def test_benchmark_result_extract_invalid_names():
br = BenchmarkResult(["loop1", "loop2"], 5, ["mae", "mse"])
with pytest.raises(KeyError):
br.extract_metric_as_array("invalid loop", "mse")
with pytest.raises(KeyError):
br.extract_metric_as_array("loop1", "invalid metric")
def benchmark_result():
br = BenchmarkResult(["loop1", "loop2"], 2, ["mae", "mse"])
br.add_results("loop1", 0, "mae", np.array([1, 2, 3]))
br.add_results("loop1", 0, "mse", np.array([1, 2]))
br.add_results("loop1", 1, "mae", np.array([4, 5, 6]))
br.add_results("loop1", 1, "mse", np.array([1, 2]))
br.add_results("loop2", 0, "mae", np.array([1]))
br.add_results("loop2", 0, "mse", np.array([2]))
br.add_results("loop2", 1, "mae", np.array([4]))
br.add_results("loop2", 1, "mse", np.array([1]))
return br
def test_benchmark_result_add_data():
br = BenchmarkResult(['loop1'], 2, ['mae', 'mse'])
br.add_results('loop1', 0, 'mae', np.array([1, 2, 3]))
br.add_results('loop1', 0, 'mse', np.array([1, 2]))
br.add_results('loop1', 1, 'mae', np.array([4, 5, 6]))
br.add_results('loop1', 1, 'mse', np.array([1, 2]))
mae_data = br.extract_metric_as_array('loop1', 'mae')
assert_array_equal(mae_data, np.array([[1, 2, 3], [4, 5, 6]]))
mse_data = br.extract_metric_as_array('loop1', 'mse')
assert_array_equal(mse_data, np.array([[1, 2], [1, 2]]))
def test_benchmark_result_add_data():
br = BenchmarkResult(["loop1"], 2, ["mae", "mse"])
br.add_results("loop1", 0, "mae", np.array([1, 2, 3]))
br.add_results("loop1", 0, "mse", np.array([1, 2]))
br.add_results("loop1", 1, "mae", np.array([4, 5, 6]))
br.add_results("loop1", 1, "mse", np.array([1, 2]))
mae_data = br.extract_metric_as_array("loop1", "mae")
assert_array_equal(mae_data, np.array([[1, 2, 3], [4, 5, 6]]))
mse_data = br.extract_metric_as_array("loop1", "mse")
assert_array_equal(mse_data, np.array([[1, 2], [1, 2]]))
def benchmark_result():
br = BenchmarkResult(['loop1', 'loop2'], 2, ['mae', 'mse'])
br.add_results('loop1', 0, 'mae', np.array([1, 2, 3]))
br.add_results('loop1', 0, 'mse', np.array([1, 2]))
br.add_results('loop1', 1, 'mae', np.array([4, 5, 6]))
br.add_results('loop1', 1, 'mse', np.array([1, 2]))
br.add_results('loop2', 0, 'mae', np.array([1]))
br.add_results('loop2', 0, 'mse', np.array([2]))
br.add_results('loop2', 1, 'mae', np.array([4]))
br.add_results('loop2', 1, 'mse', np.array([1]))
return br