def predict():
"""
The second step is optional and it isn't necessary, you can comment
that lines of code, but if you want to have a robust control of
the data that people post in your API you have to include in a correct
format in the same way that the example in validation.py
"""
if request.method == 'POST':
# Step 1: Extract POST data from request body as JSON
json_data = request.get_json()
_logger.debug(f'Inputs: {json_data}')
# Step 2: Validate the input using marshmallow schema
input_data, errors = validate_inputs(input_data=json_data)
# Step 3: Model prediction
result = make_prediction(input_data=input_data)
_logger.debug(f'Outputs: {result}')
# Step 4: Convert numpy ndarray to list
predictions = result.get('predictions').tolist()
version = result.get('version')
# Step 5: Return the response as JSON
return jsonify({
'predictions': predictions,
'version': version,
'errors': errors
})
def batch_output_specific_model(model_id):
"""Output of a specific model with test.csv data
parameters
---------------
model_id: str
returns
---------------
"""
model_id = model_id
auc_score = ''
# lightgbm_output_v0.1-1588759220.335498
if request.method == 'POST':
data_mngmnt = Data()
data_mngmnt.from_csv("test.csv", sep=',')
data = data_mngmnt.df
result = make_prediction(input_data=data, id_model=model_id)
predictions = result.get('predictions').tolist()
if 'default' in data.columns:
auc_score = get_auc_score(data['default'],
result.get('predictions'))
return jsonify({
'result': predictions,
'model': model_id,
'auc_score': auc_score
})
def test_make_single_prediction():
# Given
test_data = load_dataset(file_name='test.csv')
single_test_json = test_data[0:1]
# When
subject = make_prediction(input_data=single_test_json)
# Then
assert subject is not None
assert isinstance(subject.get('predictions')[0], float)
assert math.ceil(subject.get('predictions')[0]) == 112476
def test_make_multiple_predictions():
# Given
test_data = load_dataset(file_name='test.csv')
original_data_length = len(test_data)
multiple_test_json = test_data
# When
subject = make_prediction(input_data=multiple_test_json)
# Then
assert subject is not None
assert len(subject.get('predictions')) == 1451
# We expect some rows to be filtered out
assert len(subject.get('predictions')) != original_data_length
def outputs_upload(model_id):
"""Upload a csv file and return prediction output
"""
if request.method == 'POST':
auc_score = ''
# Create variable for uploaded file
df = pd.read_csv(request.files.get('fileupload'))
result = make_prediction(input_data=df, id_model=model_id)
predictions = result.get('predictions').tolist()
if 'default' in df.columns:
auc_score = get_auc_score(df['default'], result.get('predictions'))
return jsonify({
'result': predictions,
'model': model_id,
'auc_score': auc_score
})
def predict():
if request.method == "POST":
# Step 1: Extract POST data from request body as JSON
json_data = request.get_json()
# Step 2: Access the model prediction function (also validates data)
result = make_prediction(input_data=json_data)
# Step 3: Handle errors
errors = result.get("errors")
if errors:
return Response(json.dumps(errors), status=400)
# Step 4: Split out results
predictions = result.get("predictions").tolist()
version = result.get("version")
# Step 5: Prepare prediction response
return jsonify({
"predictions": predictions,
"version": version,
"errors": errors
})
def output_specific_model(model_id):
model_id = model_id
# lightgbm_output_v0.1-1588759220.335498
if request.method == 'POST':
# Step 1: Extract POST data from request body as JSON
data = request.get_json()
# Step 2: Validate the input using marshmallow schema
data, errors = validate_inputs(input_data=data)
# Step 3: Model prediction
result = make_prediction(input_data=data, id_model=model_id)
predictions = result.get('predictions').tolist()
version = result.get('version')
return jsonify({
'result': predictions,
'model': model_id,
'version': version,
'errors': errors
})
def predict():
"""Output of default model based on last timestamp"""
if request.method == 'POST':
# Step 1: Extract POST data from request body as JSON
json_data = request.get_json()
# _logger.debug(f'Inputs: {json_data}')
# Step 2: Validate the input using marshmallow schema
input_data, errors = validate_inputs(input_data=json_data)
# Step 3: Model prediction
result = make_prediction(input_data=input_data)
_logger.debug(f'Outputs: {result}')
# Step 4: Convert numpy ndarray to list
predictions = result.get('predictions').tolist()
version = result.get('version')
# Step 5: Return the response as JSON
return jsonify({
'predictions': predictions,
'version': version,
'errors': errors
})