larry-discuss team mailing list archive

[Keith Goodman <kwgoodman@xxxxxxxxx>]

On Thu, Feb 4, 2010 at 11:55 AM,  <josef.pktd@xxxxxxxxx> wrote:
> On Thu, Feb 4, 2010 at 2:42 PM,  <josef.pktd@xxxxxxxxx> wrote:
>> I just ran the test coverage with figleaf  (copied from html so a bit
>> messy formatting)
>>
>> Dependencies: figleaf
>> Filename        # lines # covered       % covered
>> totals:         1470    1052    71.6%
>> C:\Josef\eclipsegworkspace\larrylp\larrylp\la\util\__init__.py  1       1       100.0
>> c:\josef\eclipsegworkspace\larrylp\larrylp\la\version.py        2       2       100.0
>> C:\Josef\eclipsegworkspace\larrylp\larrylp\la\util\scipy.py     74      72      97.3
>> c:\josef\eclipsegworkspace\larrylp\larrylp\la\afunc.py  233     215     92.3
>> c:\josef\eclipsegworkspace\larrylp\larrylp\la\__init__.py       16      14      87.5
>> c:\josef\eclipsegworkspace\larrylp\larrylp\la\deflarry.py       813     690     84.9
>> c:\josef\eclipsegworkspace\larrylp\larrylp\la\util\misc.py      67      49      73.1
>> c:\josef\eclipsegworkspace\larrylp\larrylp\la\func.py   42      6       14.3
>> c:\josef\eclipsegworkspace\larrylp\larrylp\la\io.py     222     3       1.4
>>
>> Josef
>>
>
> summary
>
> deflarry
> larry methods without tests: keep_label, keep_x, and the new function
> totuples, tolist, fromtuples, fromlist, todict, fromdict
> there might be some other one-liner methods
> the rest are mostly exceptions that are never reached
>
> afunc has only ranking with norm Gaussian that is not tested, the rest
> are uncovered exceptions
>
> func has essentially zero coverage
>
> util misc looks like full coverage except for exceptions
>
> coverage of io is incorrect because I don't have compatible h5py

These look like good places to try out the new assert_larry function.