Fixed #297 update tests to check error strings

[jeff-hernandez]

• On windows platform, there is an open issue currently in pandas where it raises an error when reading a file with accents in the file path (i.e. régions.csv). So, I resolved it with the following:

# featuretools\tests\testing_utils\mock_ds.py:334
df = pd.read_csv(open(filenames[entity], 'r', encoding='utf8'), encoding='utf-8')

• This snippet np.dtype((np.integer, np.floating)).type was causing this issue. So, I resolved it by changing it to the following:

np.issubdtype(time, np.integer) or np.issubdtype(time, np.floating)

• Not sure how to get the error text for test_not_enough_memory

[CLAassistant]

All committers have signed the CLA.

[kmax12]

thanks @jeff-hernandez. circleci is having problems sending the status back to github right now, but looks like one test is failing: https://circleci.com/gh/Featuretools/featuretools/1418.

the error is AssertionError: Pattern 'cutoff_time times must be datetime type.*try casting via pd.to_datetime.*' not found in 'cutoff_time times must be numeric: try casting via pd.to_numeric(cutoff_time['time'])'

[kmax12]

no need to have match text here. In this case NotEnoughData is already specific enough

[jeff-hernandez]

Thanks @kmax12. I updated the commit. Let me know if anything else needs to be changed.

[kmax12]

@jeff-hernandez looks like python 2.7 tests are failing because encoding isn't a valid parameter to open in python 2.7.

[codecov]

Codecov Report

Merging #303 into master will increase coverage by 0.02%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master     #303      +/-   ##
==========================================
+ Coverage   95.03%   95.06%   +0.02%     
==========================================
  Files          71       71              
  Lines        7620     7660      +40     
==========================================
+ Hits         7242     7282      +40     
  Misses        378      378

Impacted Files	Coverage Δ
featuretools/tests/testing_utils/mock_ds.py	87.4% <ø> (ø)	⬆️
...uretools/tests/entityset_tests/test_es_metadata.py	96.77% <100%> (+0.07%)	⬆️
...aturetools/tests/entityset_tests/test_timedelta.py	100% <100%> (ø)	⬆️
featuretools/tests/entityset_tests/test_es.py	99.37% <100%> (+0.01%)	⬆️
featuretools/tests/entityset_tests/test_entity.py	100% <100%> (ø)	⬆️
...utational_backend/test_calculate_feature_matrix.py	99.3% <100%> (+0.01%)	⬆️
...ols/tests/feature_function_tests/test_agg_feats.py	98.56% <100%> (+0.01%)	⬆️
featuretools/utils/wrangle.py	79.77% <100%> (ø)	⬆️
...ests/computational_backend/test_encode_features.py	100% <100%> (ø)	⬆️
.../feature_function_tests/test_transform_features.py	98.85% <100%> (ø)	⬆️
... and 1 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 449df54...a027e19. Read the comment docs.

[jeff-hernandez]

I switched back to the previous snippet:

df = pd.read_csv(filenames[entity], encoding='utf-8')

I still get the error for reading a file with accents in the file path. I couldn't get the error text to work in the function test_add_relationship_errors_on_dtype_mismatch because of unicode related issues.

[kmax12]

does the following work for test_add_relationship_errors_on_dtype_mismatch to handle unicode?

error_text = u'Unable to add relationship because id in régions is Pandas dtype object and session_id in log2 is Pandas dtype int64.'
with pytest.raises(ValueError, match=error_text):
    mismatch = Relationship(entityset[u'régions']['id'], entityset['log2']['session_id'])
    entityset.add_relationship(mismatch)

other than that, it's good to merge!

[jeff-hernandez]

@kmax12 I tried the code snippet. This is the error I get for python 2.7:

UnicodeEncodeError: 'ascii' codec can't encode character u'\xe9' in position 42: ordinal not in range(128)

[kmax12]

@jeff-hernandez im looking into this now. weirdly it's only a problem in python 2.7.

[kmax12]

actually, here's an easier fix. let's just change the entity for this test case. I think that this solution is fine since we aren't trying to test unicode stuff with this test case.

So, it can just be

error_text = u'Unable to add relationship because id in customers is Pandas dtype category and session_id in log2 is Pandas dtype int64.'
with pytest.raises(ValueError, match=error_text):
    mismatch = Relationship(entityset[u'customers']['id'], entityset['log2']['session_id'])
    entityset.add_relationship(mismatch)

[jeff-hernandez]

Okay sounds good. I made the adjustment.

[kmax12]

this is look good. actually one more stylistic change. In many places, you removed a blank line before the error test. can you add those back in?

for example

after that, we're good to merge

[jeff-hernandez]

No problem. I updated the adjustment. Let me know if anything else needs modification. Thanks.

[kmax12]

Looks good. Merging. Thanks for the contribution!