def Predict():
request = extract_message()
sanity_check_request(request)
features = get_data_from_json(request)
names = request.get("data",{}).get("names")
predictions = predict(user_model,features,names)
# If predictions is an numpy array or we used the default data then return as numpy array
if isinstance(predictions, np.ndarray) or "data" in request:
predictions = np.array(predictions)
if len(predictions.shape)>1:
class_names = get_class_names(user_model, predictions.shape[1])
else:
class_names = []
data = array_to_rest_datadef(predictions, class_names, request.get("data",{}))
response = {"data":data,"meta":{}}
else:
response = {"binData":predictions,"meta":{}}
tags = get_custom_tags(user_model)
if tags:
response["meta"]["tags"] = tags
metrics = get_custom_metrics(user_model)
if metrics:
response["meta"]["metrics"] = metrics
return jsonify(response)
def Route():
request = extract_message()
if debug:
print("SELDON DEBUGGING")
print("Request received: ")
print(request)
sanity_check_request(request)
datadef = request.get("data")
features = rest_datadef_to_array(datadef)
routing = np.array(
[[route(user_router, features, datadef.get("names"))]])
# TODO: check that predictions is 2 dimensional
class_names = []
data = array_to_rest_datadef(routing, class_names, datadef)
response = {"data": data, "meta": {}}
tags = get_custom_tags(user_router)
if tags:
response["meta"]["tags"] = tags
metrics = get_custom_metrics(user_router)
if metrics:
response["meta"]["metrics"] = metrics
return jsonify(response)
def SendFeedback():
feedback = extract_message()
if debug:
print("SELDON DEBUGGING")
print("Feedback received: ")
print(feedback)
datadef_request = feedback.get("request", {}).get("data", {})
features = rest_datadef_to_array(datadef_request)
datadef_truth = feedback.get("truth", {}).get("data", {})
truth = rest_datadef_to_array(datadef_truth)
reward = feedback.get("reward")
try:
routing = feedback.get("response").get("meta").get("routing").get(
PRED_UNIT_ID)
except AttributeError:
raise SeldonMicroserviceException(
"Router feedback must contain a routing dictionary in the response metadata"
)
send_feedback(user_router, features, datadef_request.get("names"),
routing, reward, truth)
return jsonify({})
def TransformInput():
request = extract_message()
sanity_check_request(request)
features = get_data_from_json(request)
names = request.get("data", {}).get("names")
transformed = transform_input(user_model, features, names)
# If predictions is an numpy array or we used the default data then return as numpy array
if isinstance(transformed, np.ndarray) or "data" in request:
new_feature_names = get_feature_names(user_model, names)
transformed = np.array(transformed)
data = array_to_rest_datadef(transformed, new_feature_names,
request.get("data", {}))
response = {"data": data, "meta": {}}
else:
response = {"binData": transformed, "meta": {}}
tags = get_custom_tags(user_model)
if tags:
response["meta"]["tags"] = tags
metrics = get_custom_metrics(user_model)
if metrics:
response["meta"]["metrics"] = metrics
return jsonify(response)
def SendFeedback():
feedback = extract_message()
datadef_request = feedback.get("request").get("data")
features = rest_datadef_to_array(datadef)
truth = rest_datadef_to_array(feedback.get("truth"))
reward = feedback.get("reward")
send_feedback(user_model,features,datadef_request.get("names"),truth,reward)
return jsonify({})
def SendFeedback():
feedback = extract_message()
datadef_request = feedback.get("request").get("data")
features = rest_datadef_to_array(datadef)
truth = rest_datadef_to_array(feedback.get("truth"))
reward = feedback.get("reward")
routing = feedback.get("response").get("meta").get("routing").get(PRED_UNIT_ID)
send_feedback(user_router,features,datadef_request.get("names"),routing,reward,truth)
return jsonify({})
def Predict():
request = extract_message()
sanity_check_request(request)
datadef = request.get("data")
features = rest_datadef_to_array(datadef)
predictions = np.array(predict(user_model,features,datadef.get("names")))
# TODO: check that predictions is 2 dimensional
class_names = get_class_names(user_model, predictions.shape[1])
data = array_to_rest_datadef(predictions, class_names, datadef)
return jsonify({"data":data})
def TransformInput():
request = extract_message()
sanity_check_request(request)
datadef = request.get("data")
features = rest_datadef_to_array(datadef)
outlier_scores = score(user_model,features,datadef.get("names"))
# TODO: check that predictions is 2 dimensional
request["meta"].setdefault("tags",{})
request["meta"]["tags"]["outlierScore"] = list(outlier_scores)
return jsonify(request)
def TransformOutput():
request = extract_message()
sanity_check_request(request)
datadef = request.get("data")
features = rest_datadef_to_array(datadef)
transformed = np.array(transform_output(user_model,features,datadef.get("names")))
# TODO: check that predictions is 2 dimensional
new_class_names = get_class_names(user_model, datadef.get("names"))
data = array_to_rest_datadef(transformed, new_class_names, datadef)
return jsonify({"data":data})
def Route():
request = extract_message()
sanity_check_request(request)
datadef = request.get("data")
features = rest_datadef_to_array(datadef)
routing = np.array([[route(user_router,features,datadef.get("names"))]])
# TODO: check that predictions is 2 dimensional
class_names = []
data = array_to_rest_datadef(routing, class_names, datadef)
return jsonify({"data":data})
def Predict():
request = extract_message()
sanity_check_request(request)
datadef = request.get("data")
features = rest_datadef_to_array(datadef)
predictions = np.array(
predict(user_model, features, datadef.get("names")))
if len(predictions.shape) > 1:
class_names = get_class_names(user_model, predictions.shape[1])
else:
class_names = []
data = array_to_rest_datadef(predictions, class_names, datadef)
return jsonify({"data": data})
def Predict():
request = extract_message()
sanity_check_request(request)
datadef = request.get("data")
features = rest_datadef_to_array(datadef)
predictions = np.array(predict(user_model,features,datadef.get("names")))
if len(predictions.shape)>1:
class_names = get_class_names(user_model, predictions.shape[1])
else:
class_names = []
data = array_to_rest_datadef(predictions, class_names, datadef)
response = {"data":data,"meta":{}}
tags = get_custom_tags(user_model)
if tags:
response["meta"]["tags"] = tags
metrics = get_custom_metrics(user_model)
if metrics:
response["meta"]["metrics"] = metrics
return jsonify(response)
def TransformOutput():
request = extract_message()
sanity_check_request(request)
features = get_data_from_json(request)
names = request.get("data", {}).get("names")
transformed = transform_output(user_model, features, names)
if isinstance(transformed, np.ndarray) or "data" in request:
new_class_names = get_class_names(user_model, names)
data = array_to_rest_datadef(transformed, new_class_names,
request.get("data", {}))
response = {"data": data, "meta": {}}
else:
response = {"binData": transformed, "meta": {}}
tags = get_custom_tags(user_model)
if tags:
response["meta"]["tags"] = tags
metrics = get_custom_metrics(user_model)
if metrics:
response["meta"]["metrics"] = metrics
return jsonify(response)
def Predict():
request = extract_message()
sanity_check_request(request)
datadef = request.get("data")
return jsonify(predict(user_model, request, datadef.get("names")))
