All committers have signed the CLA.

Is the fact that it failed with Python 2.7 actually an issue?

Hi @samuel500, What's the idea behind this pull requests?
Travis fails because of pylint issues

** Module watson_developer_cloud.natural_language_understanding_v1
C: 39, 0: No space allowed around keyword argument assignment
            x_watson_learning_opt_out = x_watson_learning_opt_out)
                                      ^ (bad-whitespace)

Once you fix the whitespace issue run sh pylint.sh to verify that the code passes the code style rules

Codecov Report

Merging #246 into master will decrease coverage by 2.87%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #246      +/-   ##
==========================================
- Coverage   84.77%   81.89%   -2.88%     
==========================================
  Files          24       24              
  Lines        1182     1182              
==========================================
- Hits         1002      968      -34     
- Misses        180      214      +34

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 33a653b...b6f9a56. Read the comment docs.

Ok, fixed the pylint issue.
The idea behind the pull request is to make it possible to opt out of Watson learning while using natural language understanding. For the service I'm building I need to give the option to users that the data they provide will not be used in any way. That is not possible right now using natural_language_understanding_v1.py.

Here's the documentation that is inaccurate without my pull:
https://www.ibm.com/watson/developercloud/natural-language-understanding/api/v1/?python#data-collection