Skip to content

Commit

Permalink
code review suggestions #2
Browse files Browse the repository at this point in the history
  • Loading branch information
@RafalSkolasinski @seldondev
RafalSkolasinski authored and seldondev committed Sep 18, 2020
1 parent 31a5fad commit 09046d8
Show file tree
Hide file tree
Showing 3 changed files with 189 additions and 139 deletions.
149 changes: 101 additions & 48 deletions python/seldon_core/seldon_methods.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -106,31 +106,49 @@ def predict(
if is_proto:
(features, meta, datadef, data_type) = extract_request_parts(request)

client_response, runtime_metrics, runtime_tags = client_predict(
client_response = client_predict(
user_model, features, datadef.names, meta=meta
)

metrics = client_custom_metrics(
user_model, seldon_metrics, PREDICT_METRIC_METHOD_TAG, runtime_metrics,
user_model,
seldon_metrics,
PREDICT_METRIC_METHOD_TAG,
client_response.metrics,
)

return construct_response(
user_model, False, request, client_response, meta, metrics, runtime_tags
user_model,
False,
request,
client_response.data,
meta,
metrics,
client_response.tags,
)
else:
(features, meta, datadef, data_type) = extract_request_parts_json(request)
class_names = datadef["names"] if datadef and "names" in datadef else []

client_response, runtime_metrics, runtime_tags = client_predict(
client_response = client_predict(
user_model, features, class_names, meta=meta
)

metrics = client_custom_metrics(
user_model, seldon_metrics, PREDICT_METRIC_METHOD_TAG, runtime_metrics,
user_model,
seldon_metrics,
PREDICT_METRIC_METHOD_TAG,
client_response.metrics,
)

return construct_response_json(
user_model, False, request, client_response, meta, metrics, runtime_tags
user_model,
False,
request,
client_response.data,
meta,
metrics,
client_response.tags,
)


Expand Down Expand Up @@ -183,27 +201,30 @@ def send_feedback(
)
routing = request.response.meta.routing.get(predictive_unit_id)

client_response, runtime_metrics, runtime_tags = client_send_feedback(
client_response = client_send_feedback(
user_model, features, datadef_request.names, reward, truth, routing
)

metrics = client_custom_metrics(
user_model, seldon_metrics, FEEDBACK_METRIC_METHOD_TAG, runtime_metrics,
user_model,
seldon_metrics,
FEEDBACK_METRIC_METHOD_TAG,
client_response.metrics,
)

if client_response is None:
client_response = np.array([])
if client_response.data is None:
client_response.data = np.array([])
else:
client_response = np.array(client_response)
client_response.data = np.array(client_response.data)

return construct_response(
user_model,
False,
request.request,
client_response,
client_response.data,
None,
metrics,
runtime_tags,
client_response.tags,
)


Expand Down Expand Up @@ -255,37 +276,49 @@ def transform_input(
if is_proto:
(features, meta, datadef, data_type) = extract_request_parts(request)

client_response, runtime_metrics, runtime_tags = client_transform_input(
client_response = client_transform_input(
user_model, features, datadef.names, meta=meta
)

metrics = client_custom_metrics(
user_model,
seldon_metrics,
INPUT_TRANSFORM_METRIC_METHOD_TAG,
runtime_metrics,
client_response.metrics,
)

return construct_response(
user_model, False, request, client_response, meta, metrics, runtime_tags
user_model,
False,
request,
client_response.data,
meta,
metrics,
client_response.tags,
)
else:
(features, meta, datadef, data_type) = extract_request_parts_json(request)
class_names = datadef["names"] if datadef and "names" in datadef else []

client_response, runtime_metrics, runtime_tags = client_transform_input(
client_response = client_transform_input(
user_model, features, class_names, meta=meta
)

metrics = client_custom_metrics(
user_model,
seldon_metrics,
INPUT_TRANSFORM_METRIC_METHOD_TAG,
runtime_metrics,
client_response.metrics,
)

return construct_response_json(
user_model, False, request, client_response, meta, metrics, runtime_tags
user_model,
False,
request,
client_response.data,
meta,
metrics,
client_response.tags,
)


Expand Down Expand Up @@ -337,43 +370,49 @@ def transform_output(
if is_proto:
(features, meta, datadef, data_type) = extract_request_parts(request)

client_response, runtime_metrics, runtime_tags = client_transform_output(
client_response = client_transform_output(
user_model, features, datadef.names, meta=meta
)

metrics = client_custom_metrics(
user_model,
seldon_metrics,
OUTPUT_TRANSFORM_METRIC_METHOD_TAG,
runtime_metrics,
client_response.metrics,
)

return construct_response(
user_model,
False,
request,
client_response,
client_response.data,
meta,
metrics,
runtime_tags,
client_response.tags,
)
else:
(features, meta, datadef, data_type) = extract_request_parts_json(request)
class_names = datadef["names"] if datadef and "names" in datadef else []

client_response, runtime_metrics, runtime_tags = client_transform_output(
client_response = client_transform_output(
user_model, features, class_names, meta=meta
)

metrics = client_custom_metrics(
user_model,
seldon_metrics,
OUTPUT_TRANSFORM_METRIC_METHOD_TAG,
runtime_metrics,
client_response.metrics,
)

return construct_response_json(
user_model, False, request, client_response, meta, metrics, runtime_tags
user_model,
False,
request,
client_response.data,
meta,
metrics,
client_response.tags,
)


Expand Down Expand Up @@ -418,35 +457,53 @@ def route(
client_response = client_route(
user_model, features, datadef.names, meta=meta
)
if not isinstance(client_response, int):
if not isinstance(client_response.data, int):
raise SeldonMicroserviceException(
"Routing response must be int but got " + str(client_response)
"Routing response must be int but got " + str(client_response.data)
)
client_response_arr = np.array([[client_response]])
client_response_arr = np.array([[client_response.data]])

metrics = client_custom_metrics(
user_model, seldon_metrics, ROUTER_METRIC_METHOD_TAG
user_model,
seldon_metrics,
ROUTER_METRIC_METHOD_TAG,
client_response.metrics,
)

return construct_response(
user_model, False, request, client_response_arr, None, metrics
user_model,
False,
request,
client_response_arr,
None,
metrics,
client_response.tags,
)
else:
(features, meta, datadef, data_type) = extract_request_parts_json(request)
class_names = datadef["names"] if datadef and "names" in datadef else []
client_response = client_route(user_model, features, class_names, meta=meta)
if not isinstance(client_response, int):
if not isinstance(client_response.data, int):
raise SeldonMicroserviceException(
"Routing response must be int but got " + str(client_response)
"Routing response must be int but got " + str(client_response.data)
)
client_response_arr = np.array([[client_response]])
client_response_arr = np.array([[client_response.data]])

metrics = client_custom_metrics(
user_model, seldon_metrics, ROUTER_METRIC_METHOD_TAG
user_model,
seldon_metrics,
ROUTER_METRIC_METHOD_TAG,
client_response.metrics,
)

return construct_response_json(
user_model, False, request, client_response_arr, None, metrics
user_model,
False,
request,
client_response_arr,
None,
metrics,
client_response.tags,
)


Expand Down Expand Up @@ -518,25 +575,23 @@ def merge_metrics(meta_list, custom_metrics):
names_list.append(datadef.names)
meta_list.append(meta)

client_response, runtime_metrics, runtime_tags = client_aggregate(
user_model, features_list, names_list
)
client_response = client_aggregate(user_model, features_list, names_list)

metrics = client_custom_metrics(
user_model,
seldon_metrics,
AGGREGATE_METRIC_METHOD_TAG,
runtime_metrics,
client_response.metrics,
)

return construct_response(
user_model,
False,
request.seldonMessages[0],
client_response,
client_response.data,
merge_meta(meta_list),
merge_metrics(meta_list, metrics),
runtime_tags,
client_response.tags,
)
else:
features_list = []
Expand All @@ -561,25 +616,23 @@ def merge_metrics(meta_list, custom_metrics):
names_list.append(class_names)
meta_list.append(meta)

client_response, runtime_metrics, runtime_tags = client_aggregate(
user_model, features_list, names_list
)
client_response = client_aggregate(user_model, features_list, names_list)

metrics = client_custom_metrics(
user_model,
seldon_metrics,
AGGREGATE_METRIC_METHOD_TAG,
runtime_metrics,
client_response.metrics,
)

return construct_response_json(
user_model,
False,
msgs[0],
client_response,
client_response.data,
merge_meta(meta_list),
merge_metrics(meta_list, metrics),
runtime_tags,
client_response.tags,
)


Expand Down
Loading

0 comments on commit 09046d8

Please sign in to comment.