def main(args):
if args.tool == 'test':
test(args)
exit(0)
if args.tool == 'validate':
validate(args)
exit(0)
if args.tool == 'convert':
convert(args)
exit(0)
print('No such tool \'{}\''.format(args.tool))
exit(1)
def test_TemplateData_with_extra_translations_dont_validate(self):
with self.assertRaises(Exception):
validate({
"TemplateData": {
"example": {
"args": [],
"translations": {
"arg1": "someValue",
},
},
},
"RunList": [],
})
def test_RunList_with_extra_annotations_fail_to_validate(self):
with self.assertRaises(Exception):
validate({
"TemplateData": {
"example": {
"args": [
"arg1",
],
"annotations": {
"arg1": "someValue",
"extraAnnotation": "someValue",
},
},
},
"RunList": []
})
def test_RunList_with_extra_args_fail_to_validate(self):
with self.assertRaises(Exception):
validate({
"TemplateData": {
"example": {
"args": [],
"translations": {
"arg1": "someValue",
},
},
},
"RunList": [{
"template_type": "example",
"args": {
"a": "b",
"arg1": 5,
},
}],
})
def test_runs_have_default_tags(self):
v = validate({
"TemplateData": {
"HBIR": self.sample_hbir_template,
},
"RunList": [{
"template_type": "HBIR",
"args": self.sample_args,
}]
})
self.assertTrue(v["RunList"][0]["tag"])
def test_there_should_be_TemplateData_if_there_are_RunList(self):
with self.assertRaises(Exception):
self.assertFalse(
validate({
"TemplateData": {},
"RunList": [{
"template_type": "NONE",
"args": {
"a": "b"
},
}],
}))
def evaluate_predictive(self, test_x, test_y):
"""
Takes two numpy arrays test_x, test_y:
returns a dictionary with the metrics
to evaluate a predictive model (without comebacks)
and also the training time + inference time
"""
print("=== Evaluating predictive model...")
#test_x = self._preprocessor.preprocess(test_x)
pred = self._model.predict(test_x)
metrics = validate(test_y, pred)
metrics['train_time'] = self.train_time
return metrics
def test_RunList_with_ommitted_with_no_defaults_fail_to_validate(self):
with self.assertRaises(Exception):
validate({
"TemplateData": {
"example": {
"args": [
"arg1",
"noDefaults",
],
"translations": {
"arg1": "someValue",
},
"prop_options": {
"arg1": "defaultvalue",
},
},
},
"RunList": [{
"template_type": "example",
"args": {
"arg1": 5,
},
}],
})
def test_runs_have_custom_tags(self):
custom_tag = random_run_id()
v = validate({
"TemplateData": {
"HBIR": self.sample_hbir_template,
},
"RunList": [{
"template_type": "HBIR",
"args": self.sample_args,
"tag": custom_tag,
}]
})
self.assertEqual(v["RunList"][0]["tag"], custom_tag)
def evaluate(self, test_x, test_y):
"""
Takes two numpy arrays test_x, test_y:
returns a dictionary with the metrics
to evaluate a productive model (with comebacks)
and also the training time + inference time
"""
self.test_size = test_x.shape[0]
print('{:=^80}'.format(' VALIDATE MODEL '))
print("Evaluating productive model...")
pred_y = self._infer(test_x)
# Validamos las predicciones de test
metrics = validate(test_y, pred_y)
# Incluimos metricas de training
metrics['train_time'] = self.train_time
metrics['mean_inference_time'] = self.inference_time / self.test_size
metrics['train_size'] = self.train_size
metrics['test_size'] = self.test_size
print('{:=^80}'.format(''))
return metrics
def do_validate(arguments):
"""
Validate a configuration based on the schema provided.
"""
with open(arguments['<config>'], 'r') as f:
args = json.loads(f.read())
try:
if validate.validate_blackbox(args) is not None:
return
except Exception as e:
print(e)
print("attempting to validate SPECtate configuration...")
try:
return validate.validate(args) is None
except Exception as e:
return e
return True
def test_RunList_with_times_validates(self):
v = validate({
"TemplateData": {
"HBIR": self.sample_hbir_template,
},
"RunList": [
{
"template_type": "HBIR",
"args": self.sample_args,
},
{
"template_type": "HBIR",
"args": self.sample_args,
"times": 2,
},
]
})
for run in v["RunList"]:
self.assertEqual(self.sample_args, run["args"])
self.assertEqual(v["RunList"][0]["times"], 1)
self.assertEqual(v["RunList"][1]["times"], 2)
def test_data_set_two_true_scenario():
string_column_names = ["c"]
assert validate(data_set_two, string_column_names)
def test_data_set_two_false_scenario_two():
string_column_names = []
assert not validate(data_set_two, string_column_names)
def test_data_set_three_true_scenario_two():
string_column_names = ["d"]
assert validate(data_set_three, string_column_names)
def test_invalid_config_does_not_validate(self):
with self.assertRaises(Exception):
with open('tests/fixtures/sample_config_invalid.json') as f:
j = json.loads(f.read())
self.assertFalse(validate(j))
def test_data_set_one_true_scenario():
string_column_names = ["a", "b", "c"]
assert validate(data_set_one, string_column_names)
def test_example_config_does_validate(self):
for example in self.examples:
with open(example) as f:
j = json.loads(f.read())
self.assertTrue(validate(j))
sys.path.append(PROJECT_DIR)
from src import bracket as src_bracket
from src import compute as src_compute
from src import validate as src_validate
INPUT_EXPRESSION = '请输入要计算的表达式，退出程序请输q：'
RESULT = '计算结果：{0}'
# 使用 __name__ 的目的：
# 只有执行 python index.py 时，以下代码才执行
# 如果其他人导入该模块，以下代码不执行
if __name__ == "__main__":
expression = ''
while expression != 'q':
expression = input(INPUT_EXPRESSION)
if expression == 'q':
exit()
else:
# 如果输入的表达式合法
if src_validate.validate(expression):
# 调用remove_bracket函数去除所有括号,重新赋值给expression
# 下面一行用于测试
expression = src_bracket.remove_bracket(expression)
#print('去掉了全部括号！当前表达式为：',expression)
# 调用compute函数进行加减乘除运算
result = src_compute.compute(expression)
# 如果result是小数且小数点后字符为0，就去掉小数点及其之后的字符，即按整型打印
result = re.sub('\.0', '', result)
print(RESULT.format(result))
def test_data_set_three_false_scenario():
string_column_names = ["a"]
assert not validate(data_set_three, string_column_names)