Skip to content

Commit

Permalink
Merge pull request #3271 from ryandawsonuk/3270-unit-test-req-logger
Browse files Browse the repository at this point in the history 
some unit tests for req logger
  • Loading branch information
@seldondev
seldondev authored Jun 7, 2021
2 parents b87004d + 01f6123 commit 547d083
Show file tree
Hide file tree
Showing 2 changed files with 90 additions and 2 deletions.
4 changes: 2 additions & 2 deletions components/seldon-request-logger/app/default_logger.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -475,7 +475,7 @@ def createElementsWithMetadata(X, names, results, metadata_schema, message_type)
#we want field names from metadata if available - also build a dict of metadata by name for easy lookup
metadata_dict = {}

if metadata_schema['requests'] is not None and message_type == "request":
if 'requests' in metadata_schema and message_type == "request":
names = []
#we'll get names from metadata, assuming field order to match the request
for elem in metadata_schema['requests']:
Expand All @@ -491,7 +491,7 @@ def createElementsWithMetadata(X, names, results, metadata_schema, message_type)
names.append(elem['name'])
metadata_dict[elem['name']] = elem

if metadata_schema['responses'] is not None and message_type == "response":
if 'responses' in metadata_schema and message_type == "response":
names = []
#we'll get names from metadata, assuming field order to match the response
for elem in metadata_schema['responses']:
Expand Down
88 changes: 88 additions & 0 deletions components/seldon-request-logger/app/tests.py
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,88 @@
import unittest

import default_logger
import numpy as np

#for more local testing see README
class TestRequestLogger(unittest.TestCase):

def test_enriched_elements_request(self):
"""
Test that an elements array can be built using metadata to enrich a request
"""

#mixture of one_hot, categorical and float
names = ['dummy_one_hot_1', 'dummy_one_hot_2', 'dummy_categorical', 'dummy_float']
X = np.array([[0.0, 1.0, 0.0, 2.54]])
results = None
metadata_schema = {'requests': [
{'name': 'dummy_one_hot', 'type': 'ONE_HOT', 'data_type': 'INT', 'n_categories': '0', 'category_map': {},
'schema': [{'name': 'dummy_one_hot_1', 'data_type': 'FLOAT'},
{'name': 'dummy_one_hot_2', 'data_type': 'FLOAT'}], 'shape': []},
{'name': 'dummy_categorical', 'type': 'CATEGORICAL', 'data_type': 'INT', 'n_categories': '2',
'category_map': {'0': 'dummy_cat_0', '1': 'dummy_cat_1'}, 'schema': [], 'shape': []},
{'name': 'dummy_float', 'type': 'REAL', 'data_type': 'FLOAT', 'n_categories': '0', 'category_map': {},
'schema': [], 'shape': []}], 'responses': [
{'name': 'dummy_proba', 'type': 'PROBA', 'data_type': 'FLOAT', 'n_categories': '0', 'category_map': {},
'schema': [{'name': 'dummy_proba_0', 'data_type': 'FLOAT'},
{'name': 'dummy_proba_1', 'data_type': 'FLOAT'}], 'shape': []},
{'name': 'dummy_float', 'type': 'REAL', 'data_type': 'FLOAT', 'n_categories': '0', 'category_map': {},
'schema': [], 'shape': []}]}
message_type = 'request'

#one_hot columns should be merged under a top-level, categorical replaced with label and float as-is
expected_results = [{'dummy_one_hot': {'dummy_one_hot_1': 0.0, 'dummy_one_hot_2': 1.0}, 'dummy_categorical': 'dummy_cat_0',
'dummy_float': 2.54}]
actual_results = default_logger.createElementsWithMetadata(X,names,results,metadata_schema,message_type)
self.assertEqual(expected_results, actual_results)

def test_enriched_elements_response(self):
"""
Test that an elements array can be built using metadata to enrich a response
"""

#mixture of proba and float
names = ['dummy_proba_0', 'dummy_proba_1', 'dummy_float']
X = np.array([[0.85388188, 0.14611812, 3.65 ]])
results = None
metadata_schema = {'requests': [
{'name': 'dummy_one_hot', 'type': 'ONE_HOT', 'data_type': 'INT', 'n_categories': '0', 'category_map': {},
'schema': [{'name': 'dummy_one_hot_1', 'data_type': 'FLOAT'},
{'name': 'dummy_one_hot_2', 'data_type': 'FLOAT'}], 'shape': []},
{'name': 'dummy_categorical', 'type': 'CATEGORICAL', 'data_type': 'INT', 'n_categories': '2',
'category_map': {'0': 'dummy_cat_0', '1': 'dummy_cat_1'}, 'schema': [], 'shape': []},
{'name': 'dummy_float', 'type': 'REAL', 'data_type': 'FLOAT', 'n_categories': '0', 'category_map': {},
'schema': [], 'shape': []}], 'responses': [
{'name': 'dummy_proba', 'type': 'PROBA', 'data_type': 'FLOAT', 'n_categories': '0', 'category_map': {},
'schema': [{'name': 'dummy_proba_0', 'data_type': 'FLOAT'},
{'name': 'dummy_proba_1', 'data_type': 'FLOAT'}], 'shape': []},
{'name': 'dummy_float', 'type': 'REAL', 'data_type': 'FLOAT', 'n_categories': '0', 'category_map': {},
'schema': [], 'shape': []}]}
message_type = 'response'

#proba columns should be merged under a top-level and float as-is
expected_results = [{'dummy_proba': {'dummy_proba_0': 0.85388188, 'dummy_proba_1': 0.14611812}, 'dummy_float': 3.65}]
actual_results = default_logger.createElementsWithMetadata(X,names,results,metadata_schema,message_type)
self.assertEqual(expected_results, actual_results)

def test_not_enriched_elements_request(self):
"""
Test that an elements array can be built even without metadata, provided names given
"""

#mixture of one_hot, categorical and float
names = ['Age', 'Workclass', 'Education', 'Marital Status', 'Occupation', 'Relationship', 'Race', 'Sex', 'Capital Gain', 'Capital Loss', 'Hours per week', 'Country']
X = np.array([[53.0, 4.0, 0.0, 2.0, 8.0, 4.0, 2.0, 0.0, 0.0, 0.0, 60.0, 9.0]])
results = None

#values should be matched to names
expected_results = [{
'Age': 53.0, 'Workclass': 4.0, 'Education': 0.0, 'Marital Status': 2.0, 'Occupation': 8.0,
'Relationship': 4.0, 'Race': 2.0, 'Sex': 0.0, 'Capital Gain': 0.0, 'Capital Loss': 0.0,
'Hours per week': 60.0, 'Country': 9.0}]
actual_results = default_logger.createElementsNoMetadata(X,names,results)
self.assertEqual(expected_results, actual_results)


if __name__ == '__main__':
unittest.main()

0 comments on commit 547d083

Please sign in to comment.