def test_6_trainNNModel(self):
logging.info("Test Case : Train NN model.")
filePath = 'testUseCase/supportdata/irisNN.pmml'
logFolder = 'testUseCase/supportdata/logs'
payload = {
"batchSize": 15,
"epoch": 10,
"stepPerEpoch": 10,
"learningRate": 0.001,
"loss": "categorical_crossentropy",
"metrics": ["accuracy"],
"optimizer": "Adam",
"testSize": 0.3,
"scriptOutput": "NA",
"problemType": "classification",
"filePath": os.path.abspath(filePath),
"tensorboardLogFolder": os.path.abspath(logFolder),
"tensorboardUrl": '',
'dataFolder': ''
}
result = Training.trainNeuralNetworkModels(payload)
result = json.loads(result.__dict__['_container'][0])
self.assertEqual(result['pmmlFile'],
filePath.split('/')[-1].replace('.pmml', ''))
self.assertEqual(result['idforData'],
filePath.split('/')[-1].replace('.pmml', ''))
self.assertEqual(result['status'], 'In Progress')
self.assertEqual('pID' in result, True)
Utility.deleteTaskfromMemory(result['idforData'])
logging.info("PASSED")
def post(self,requests):
userInput=requests.body
try:
userInput=json.loads(userInput)
except:
return JsonResponse({'error':'Invalid Request Parameter'},status=400)
return Training.trainMRCNN(userInput)
def post(self, requests):
try:
requests.body
# print (requests.POST.get('filePath'))
except:
return JsonResponse({'error': 'Invalid Request Parameter'},
status=400)
return Training.trainNeuralNetworkModels(requests)
def test_4_AutoMLSendData(self):
logging.info("Test Case : Send data for AutoML.")
filePath = os.path.abspath(
'testUseCase/supportdata/mpg_data_example2.csv')
result = Training.autoMLdataprocess(filePath)
self.assertEqual(
'idforData' in json.loads(result.__dict__['_container'][0]), True)
logging.info("PASSED")
def get(self,requests):
try:
pathOffile=requests.GET['filePath']
if not pathOffile:
raise Exception("Invalid Request Parameter")
except:
return JsonResponse({'error':'Invalid Request Parameter'},status=400)
return Training.autoMLdataprocess(pathOffile)
def get(self, requests, id_for_task):
return Training.statusOfModel(id_for_task)
def test_5_AutoMLTrain(self):
logging.info(
"Test Case : Perform preprocessing and train AutoML model.")
filePath = os.path.abspath(
'testUseCase/supportdata/mpg_data_example2.csv')
result = Training.autoMLdataprocess(filePath)
tempa = json.loads(result.__dict__['_container'][0])
newPMMLFileName = 'xyz.pmml'
target_variable = 'mpg'
true = True
false = False
dataPreprocessingsteps = {
"data": [{
"position": 1,
"variable": "mpg",
"dtype": "float64",
"missing_val": 0,
"changedataType": "Continuous",
"imputation_method": "None",
"data_transformation_step": "None",
"use_for_model": true
}, {
"position": 2,
"variable": "cylinders",
"dtype": "int64",
"missing_val": 0,
"changedataType": "Continuous",
"imputation_method": "None",
"data_transformation_step": "None",
"use_for_model": true
}, {
"position": 3,
"variable": "displacement",
"dtype": "float64",
"missing_val": 0,
"changedataType": "Continuous",
"imputation_method": "None",
"data_transformation_step": "None",
"use_for_model": true
}, {
"position": 4,
"variable": "horsepower",
"dtype": "float64",
"missing_val": 6,
"changedataType": "Continuous",
"imputation_method": "Mean",
"data_transformation_step": "None",
"use_for_model": true
}, {
"position": 5,
"variable": "weight",
"dtype": "int64",
"missing_val": 0,
"changedataType": "Continuous",
"imputation_method": "None",
"data_transformation_step": "None",
"use_for_model": true
}, {
"position": 6,
"variable": "acceleration",
"dtype": "float64",
"missing_val": 0,
"changedataType": "Continuous",
"imputation_method": "None",
"data_transformation_step": "None",
"use_for_model": true
}, {
"position": 7,
"variable": "model year",
"dtype": "int64",
"missing_val": 0,
"changedataType": "Categorical",
"imputation_method": "None",
"data_transformation_step": "None",
"use_for_model": true
}, {
"position": 8,
"variable": "origin",
"dtype": "int64",
"missing_val": 0,
"changedataType": "Categorical",
"imputation_method": "None",
"data_transformation_step": "One Hot Encoding",
"use_for_model": true
}, {
"position": 9,
"variable": "car name",
"dtype": "object",
"missing_val": 0,
"changedataType": "Categorical",
"imputation_method": "None",
"data_transformation_step": "None",
"use_for_model": false
}],
"problem_type":
"Regression",
"target_variable":
target_variable,
"idforData":
tempa['idforData'],
'newPMMLFileName':
newPMMLFileName,
'filePath':
filePath,
"parameters": []
}
result2 = Training.autoMLtrainModel(dataPreprocessingsteps)
result2 = json.loads(result2.__dict__['_container'][0])
self.assertEqual('pID' in result2, True)
self.assertEqual('status' in result2, True)
self.assertEqual('newPMMLFileName' in result2, True)
self.assertEqual('targetVar' in result2, True)
self.assertEqual('problem_type' in result2, True)
self.assertEqual('idforData' in result2, True)
self.assertEqual(result2['status'], 'In Progress')
self.assertEqual(result2['newPMMLFileName'], newPMMLFileName)
self.assertEqual(result2['targetVar'], target_variable)
self.assertEqual(result2['idforData'], tempa['idforData'])
result = Utility.runningTaskList()
self.assertEqual(result.status_code, 200)
self.assertEqual(
'runningTask' in json.loads(result.__dict__['_container'][0]),
True)
self.assertEqual(
len(json.loads(result.__dict__['_container'][0])['runningTask']),
1)
idforData = tempa['idforData']
result = Training.statusOfModel(idforData)
self.assertEqual(result.status_code, 200)
result = json.loads(result.__dict__['_container'][0])
self.assertEqual('pID' in result, True)
self.assertEqual('status' in result, True)
self.assertEqual('idforData' in result, True)
idforData = tempa['idforData']
result = Utility.deleteTaskfromMemory(idforData)
self.assertEqual(
'idforData' in json.loads(result.__dict__['_container'][0]), True)
self.assertEqual(
'message' in json.loads(result.__dict__['_container'][0]), True)
self.assertEqual(
json.loads(result.__dict__['_container'][0])['idforData'],
tempa['idforData'])
self.assertEqual(
json.loads(result.__dict__['_container'][0])['message'],
'Deleted successfully')
result = Utility.runningTaskList()
self.assertEqual(result.status_code, 200)
self.assertEqual(
'runningTask' in json.loads(result.__dict__['_container'][0]),
True)
self.assertEqual(
len(json.loads(result.__dict__['_container'][0])['runningTask']),
0)
logging.info("PASSED")