def test_incomplete_submissions(self):
""" Data generation should ignore incomplete submissions
Incomplete submissions have no metrics to process.
"""
# Set as admin.
self.user.is_staff=True
self.user.save()
# Create a new submission.
submission = Submission(
user=self.user,
inventory_id=Via.inventory_id,
current_page=0
)
submission.save()
# Generate data
data = generate_data_from_sessions(self.sessions, self.user)
for submission in data['submission_counts']:
# Verify the Via count is correct
correct_count = Submission.objects.filter(
inventory_id=submission['inventory_id']
).exclude(
metric__isnull=True
).count()
self.assertEqual(submission['count'], correct_count)
def test_not_enough_user_regular(self):
""" Regular users should not be able to view statistics without data.
There must be at least 10 submissions to view.
"""
try:
# Request the data
data = generate_data_from_sessions(self.sessions, self.user)
self.fail('LookupError should have been raised. No data.')
except LookupError:
pass
def test_partial_excludes(self):
""" Data is only supposed to be generated for non staff users
for inventories that have 10 submissions. Previous tests
checked for all inventories at this boundary. Test that if
some inventories have 10 submissions and some have less,
only the ones which meet the requirement are returned
within the data set.
"""
# Create a submission for Big Five, Ambiguity, and Via
Factory.create_submission(self.user, BigFive)
Factory.create_submission(self.user, Ambiguity)
Factory.create_submission(self.user, Via)
# List of inventory id's a valid submission may have
valid_submissions = [BigFive.inventory_id, Ambiguity.inventory_id, Via.inventory_id]
# Grab the data
data = generate_data_from_sessions(self.sessions, self.user)
# A count of all submissions exists
self.assertEqual(6, len(data['submission_counts']))
# Verify the submission count is correct
for submission in data['submission_counts']:
correct_count = Submission.objects.filter(inventory_id=submission['inventory_id']).count()
self.assertEqual(submission['count'], correct_count)
# Non admin does not have analysis
self.assertNotIn('metrics_analysis', data)
# Verify that the user data is correct
for user in data['users']:
# The correct number of submissions has been attached
self.assertEqual(len(user.submissions), len(valid_submissions))
# For each user submission
for submission in user.submissions:
# The submission belongs to the user
self.assertEqual(user, submission.user)
# The submission is valid
self.assertIn(submission.inventory_id, valid_submissions)
# The correct metrics exist in the submission
num_metrics = Metric.objects.filter(submission=submission).count()
self.assertEqual(len(submission.metrics), num_metrics)
# For each submission metric
for metric in submission.metrics:
# The metric belongs to the submission
self.assertEqual(submission, metric.submission)
def test_not_enough_user_staff(self):
""" Staff users should be able to view statistics any time.
This is regardless of whether or not there are more than
10 submissions for any inventory
"""
# Make staff
self.user.is_staff = True
self.user.save()
# Request the data and verify it exists
data = generate_data_from_sessions(self.sessions, self.user)
# All three sets of subdata should exist
self.assertIn('submission_counts', data)
self.assertIn('metrics_analysis', data)
self.assertIn('users', data)
def test_enough_data(self):
""" Metrics are returned correctly if enough data exists
Independent of whether or not user is staff.
"""
# Generate the 10th set of data
Factory.create_set_of_submissions(self.user)
# Generate the data for a non staff
data = generate_data_from_sessions(self.sessions, self.user)
# Verify non staff do not have access to analysis
self.assertIn('submission_counts', data)
self.assertNotIn('metrics_analysis', data)
self.assertIn('users', data)
# Set staff to check for analysis
self.user.is_staff = True
self.user.save()
# Load the data
data = generate_data_from_sessions(self.sessions, self.user)
# Verify staff have all data
self.assertIn('submission_counts', data)
self.assertIn('metrics_analysis', data)
self.assertIn('users', data)
# Verify each submission has a submission count
self.assertEqual(6, len(data['submission_counts']))
# Verify the submission count is correct
for submission in data['submission_counts']:
correct_count = Submission.objects.filter(inventory_id=submission['inventory_id']).count()
self.assertEqual(submission['count'], correct_count)
# Verify that there is a metric analysis for each metric
# That does not belong to Via.
correct_count = Metric.objects.exclude(
submission__inventory_id=Via.inventory_id
).distinct('key', 'submission__inventory_id').count()
self.assertEqual(len(data['metrics_analysis']), correct_count)
# Verify that the metric analysis for each inventory is correct
for analysis in data['metrics_analysis']:
correct_analysis = Metric.objects.filter(
key=analysis['key'],
submission__inventory_id=analysis['submission__inventory_id']
).values(
'key', 'submission__inventory_id'
).annotate(
min=Min('value'),
max=Max('value'),
mean=Avg('value'),
standard_deviation=StdDev('value')
)
self.assertDictEqual(analysis, correct_analysis[0])
# Verify that the user data is correct
for user in data['users']:
# The correct number of submissions has been attached
num_submissions = Submission.objects.filter(user=user).count()
self.assertEqual(len(user.submissions), num_submissions)
# For each user submission
for submission in user.submissions:
# The submission belongs to the user
self.assertEqual(user, submission.user)
# The correct metrics exist in the submission
num_metrics = Metric.objects.filter(submission=submission).count()
self.assertEqual(len(submission.metrics), num_metrics)
# For each submission metric
for metric in submission.metrics:
# The metric belongs to the submission
self.assertEqual(submission, metric.submission)
