Add the phenotype module

[mgalardini]

Sorry, to have a cleaner history and avoid any mistake I preferred doing a new PR.

[codecov-io]

Current coverage is 78.77%

Merging #755 into master will increase coverage by +0.09% as of 4af4773

@@            master    #755   diff @@
======================================
  Files          315     318     +3
  Stmts        46600   47232   +632
  Branches         0       0       
  Methods          0       0       
======================================
+ Hit          36666   37209   +543
  Partial          0       0       
- Missed        9934   10023    +89

Review entire Coverage Diff as of 4af4773

Powered by Codecov. Updated on successful CI builds.

[peterjc]

Ideally that would have included a trailing comma, so that next time an entry is added to the list it is a one line change. But there's no real need to change it now.

[peterjc]

Your code includes a lot of doctest style examples (formatted with >>>), which is great. Ideally we would include these in the test suite (by adding the modules to the numpy-dependent doctest list inside run_tests.py).

Is this something you've tried?

Unfortunately making doctests testable examples can be a lot more work due to things like this and cross-platform differences, and sometimes we have to make slightly less friendly examples - or use deliberately tiny example files (e.g. see the SeqIO docstrings with for loops). That might be needed for some of your for loop examples, e.g.

    >>> for well in plate.get_row('H'):
    ...     print("%s" % well.id)
    H01
    H02
    H03
    ...

Related to this, via test_Tutorial.py we can test the doctest style examples within the LaTeX tutorial. The advantage here is we can exclude problematic lines (e.g. where we only show part of the output) within a larger example. Also @tiagoantao and @vincentdavis have been working on automatically turning the LaTeX into run-able notebooks, see e.g. http://lists.open-bio.org/pipermail/biopython-dev/2015-December/021209.html http://lists.open-bio.org/pipermail/biopython-dev/2015-December/021247.html

Note: I don't think this potential polishing should delay merging the work so far.

[peterjc]

You shouldn't use a mutable default argument, would wells=() work here? Otherwise use wells=None and set this to a (new) empty list within the __init__ method.

This is one of the problems checked for by the quantifiedcode automated checks, also pylint which calls it W0102, see e.g. http://pylint-messages.wikidot.com/messages:w0102

[mgalardini]

Ok, all the suggested changes are in (except for the doctests, which could be added later I guess?)
Thank you very much for your review!

[mgalardini]

Hi Peter,

I didn't see any comment on this PR so far; would you like me to inquiry a bit more to see if the changes are sound enough?

Best,
Marco

[mgalardini]

Hi Peter,

so far I didn't see any comments regarding this proposed module. Did you see/hear anything?

Best,
Marco

[peterjc]

I'm happy, just haven't had time outside of work to apply your changes yet. Sorry.

[mgalardini]

Hi,

importing the phenotype module will now trigger a warning.

[peterjc]

I had some issues rebasing this - which I think was due to using a Mac with its default case insensitive file system (like Windows) and git not coping well with the folder renaming.

I ended up compressing your work into one commit with some minor changes included like removing trailing white space in the tutorial and renaming another module to be lower case (Bio/phenotype/phen_micro.py). This is on a new branch which I'd like you to look at please: #828

[mgalardini]

Thanks, will have a look right now