def test_eq(self):
TestCase = namedtuple(
'TestCase', 'a b expect_eq')
test_cases = {
'equal_cases': TestCase(
a=plot.PlotData(
x=np.array([1, 2]),
y=np.array([7.46, 8.46])),
b=plot.PlotData(
x=np.array([1, 2]),
y=np.array([7.46, 8.46])),
expect_eq=True),
'not_eq': TestCase(
a=plot.PlotData(
x=np.array([1, 2]),
y=np.array([7.46, 8.46])),
b=plot.PlotData(
x=np.array([1, 2, 3]),
y=np.array([7.46, 8.46, 9.46])),
expect_eq=False),
'b_not_plot_data': TestCase(
a=plot.PlotData(
x=np.array([1, 2]),
y=np.array([7.46, 8.46])),
b={
'x': [1, 2, 3],
'y': [7.46, 8.46, 9.46],
},
expect_eq=False),
}
for test_name, test_case in test_cases.items():
with self.subTest(test_name):
is_eq = test_case.a == test_case.b
self.assertEqual(test_case.expect_eq, is_eq)
def test_build_plot_fn_raises(self):
TestCase = namedtuple(
'TestCase', 'data x_name y_name plots expected_err_type')
test_cases = {
'numeric_x_invalid_plot_specified': TestCase(
data={
"first_var = some_name": plot.PlotData(
x=np.array([1, 2]),
y=np.array([7.46, 8.46])),
},
x_name='first_var',
y_name='time',
plots="unknown",
expected_err_type=inputs.InvalidInputError),
'non_numeric_x_unsupported_plot_type_specified': TestCase(
data={
"first_var = some_name": plot.PlotData(
x=np.array(['foo', 'bar'], dtype=np.dtype('<U1')),
y=np.array([7.46, 8.46])),
},
x_name='first_var',
y_name='time',
plots=plot.AVG_LINE_TYPE,
expected_err_type=inputs.InvalidInputError),
}
for test_name, test_case in test_cases.items():
with self.subTest(test_name):
with self.assertRaises(test_case.expected_err_type):
plot_fn = plot.build_plot_fn(
test_case.data,
test_case.x_name, y_name=test_case.y_name,
plots=test_case.plots)
def test_bench_res_data(self):
TestCase = namedtuple(
'TestCase', ['split_res_items', 'expected_plot_data', 'expected_x_type', 'expected_y_type'])
test_cases = {
'int_x_float_y': TestCase(
split_res_items=[
benchmark.SplitRes(
x=1, y=7.46),
benchmark.SplitRes(
x=2, y=8.46),
],
expected_plot_data=plot.PlotData(
x=np.array([1, 2]),
y=np.array([7.46, 8.46])),
expected_x_type=np.dtype(int),
expected_y_type=np.dtype(float)),
'bool_x_float_y': TestCase(
split_res_items=[
benchmark.SplitRes(
x=True, y=7.46),
benchmark.SplitRes(
x=False, y=8.46),
],
expected_plot_data=plot.PlotData(
x=np.array([True, False]),
y=np.array([7.46, 8.46])),
expected_x_type=np.dtype(bool),
expected_y_type=np.dtype(float)),
'string_x_int_y': TestCase(
split_res_items=[
benchmark.SplitRes(
x='foo', y=1),
benchmark.SplitRes(
x='bar', y=2),
],
expected_plot_data=plot.PlotData(
x=np.array(['foo', 'bar'], dtype=np.dtype('<U1')),
y=np.array([1, 2])),
expected_x_type=np.dtype('<U1'),
expected_y_type=np.dtype(int)),
}
for test_name, test_case in test_cases.items():
with self.subTest(test_name):
plot_data = plot.bench_res_data(test_case.split_res_items)
self.assertEqual(test_case.expected_plot_data, plot_data)
self.assertEqual(test_case.expected_x_type,
plot_data.x_type())
self.assertEqual(test_case.expected_y_type,
plot_data.y_type())
def test_avg_over_x(self):
TestCase = namedtuple(
'TestCase', 'data expected_averaged')
test_cases = {
'no_work_needed': TestCase(
data=plot.PlotData(
x=np.array([1, 2]),
y=np.array([7.46, 8.46])),
expected_averaged=plot.PlotData(
x=np.array([1, 2]),
y=np.array([7.46, 8.46]))),
'3_duplicates': TestCase(
data=plot.PlotData(
x=np.array([1, 1, 1, 2, 2, 2, 3, 3, 3]),
y=np.array([1, 2, 3, 1, 2, 3, 1, 2, 3])),
expected_averaged=plot.PlotData(
x=np.array([1, 2, 3]),
y=np.array([2, 2, 2]))),
}
for test_name, test_case in test_cases.items():
with self.subTest(test_name):
avged = test_case.data.avg_over_x()
self.assertEqual(test_case.expected_averaged, avged)
def test_build_plot_fn(self):
TestCase = namedtuple(
'TestCase', 'data x_name y_name plots expected_plotfn_names')
test_cases = {
'numeric_x_no_plots_specified': TestCase(
data={
"first_var = some_name": plot.PlotData(
x=np.array([1, 2]),
y=np.array([7.46, 8.46])),
},
x_name='first_var',
y_name='time',
plots=None,
expected_plotfn_names=[plot.plot_scatter.__name__, plot.plot_avg_line.__name__]),
'numeric_x_best_fit_plot_specified': TestCase(
data={
"first_var = some_name": plot.PlotData(
x=np.array([1, 2]),
y=np.array([7.46, 8.46])),
},
x_name='first_var',
y_name='time',
plots=plot.BEST_FIT_LINE_TYPE,
expected_plotfn_names=[plot.plot_best_fit_line.__name__]),
'numeric_x_avg_bar_plot_specified': TestCase(
data={
"first_var = some_name": plot.PlotData(
x=np.array([1, 2]),
y=np.array([7.46, 8.46])),
},
x_name='first_var',
y_name='time',
plots=plot.BAR_TYPE,
expected_plotfn_names=[plot.plot_bar.__name__]),
'non_numeric_x_no_plots_specified': TestCase(
data={
"first_var = some_name": plot.PlotData(
x=np.array(['foo', 'bar'], dtype=np.dtype('<U1')),
y=np.array([7.46, 8.46])),
},
x_name='first_var',
y_name='time',
plots=None,
expected_plotfn_names=[plot.plot_bar.__name__]),
'non_numeric_x_1_plot_specified': TestCase(
data={
"first_var = some_name": plot.PlotData(
x=np.array(['foo', 'bar'], dtype=np.dtype('<U1')),
y=np.array([7.46, 8.46])),
},
x_name='first_var',
y_name='time',
plots=plot.BAR_TYPE,
expected_plotfn_names=[plot.plot_bar.__name__]),
}
for test_name, test_case in test_cases.items():
with self.subTest(test_name):
plot_fn = plot.build_plot_fn(
test_case.data,
test_case.x_name, y_name=test_case.y_name,
plots=test_case.plots)
plot_fn_names = []
if isinstance(plot_fn, list):
plot_fn_names = list(map(lambda x: x.__name__, plot_fn))
else:
plot_fn_names = [plot_fn.__name__]
self.assertEqual(
test_case.expected_plotfn_names, plot_fn_names)