QH6228 Bug fix - remove offset from coords

[jmetz]

Fixes: #871

[coveralls]

Coverage decreased (-0.57%) when pulling add4b51 on jezmpez:convex_hull_image_qhull_precision_fix into 73bf701 on scikit-image:master.

[jni]

As I recall, this approach also fixed the bug I found ~6 months ago (#672).

@jmetz, can you add your failing test case to the tests in test_convex_hull.py, and make sure they pass on your machine?

@stefanv, the decreased coverage appears to indicate that the convex hull tests are skipped on Travis. Is there a reason for this?

[jmetz]

@jni - just having a minor compile issue while trying to run my additional testing; am I missing something simple..?
python setup.py build

gives

... 
building 'skimage.filter.rank.generic_cy' extension
compiling C sources
C compiler: x86_64-linux-gnu-gcc -pthread -fno-strict-aliasing -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -fPIC

compile options: '-I/usr/local/lib/python2.7/dist-packages/numpy/core/include -I/usr/local/lib/python2.7/dist-packages/numpy/core/include -I/usr/include/python2.7 -c'
x86_64-linux-gnu-gcc: rank/generic_cy.c
x86_64-linux-gnu-gcc: error: rank/generic_cy.c: No such file or directory
x86_64-linux-gnu-gcc: fatal error: no input files
compilation terminated.
x86_64-linux-gnu-gcc: error: rank/generic_cy.c: No such file or directory
x86_64-linux-gnu-gcc: fatal error: no input files
compilation terminated.
error: Command "x86_64-linux-gnu-gcc -pthread -fno-strict-aliasing -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -fPIC -I/usr/local/lib/python2.7/dist-packages/numpy/core/include -I/usr/local/lib/python2.7/dist-packages/numpy/core/include -I/usr/include/python2.7 -c rank/generic_cy.c -o build/temp.linux-x86_64-2.7/rank/generic_cy.o" failed with exit status 4

[jmetz]

@jni - OK I worked around that by manually calling cython on most of the pyx files in skimage/filters/rank/ - did something get broken in the setup process?

Also I can confirm that my workaround also fixed the other previously failing test (test_pathological_qhull_example) .

[coveralls]

Coverage decreased (-0.57%) when pulling 920c3ef on jmetz:convex_hull_image_qhull_precision_fix into 73bf701 on scikit-image:master.

[JDWarner]

Trailing whitespace

[JDWarner]

I noticed a few very minor style issues about unnecessary whitespace; I think I noted them all. Many editors can be set to automatically trim all trailing whitespace on saves, to avoid this in the future.

Travis appears to be compiling and running the new file correctly, so I think this is ready to go once that whitespace gets trimmed!

[jmetz]

Thanks for the catch @JDWarner - I modified my vimrc to remove trailing whitespaces on save and updated the corresponding edits accordingly.
Did I get them all?

[coveralls]

Coverage decreased (-0.57%) when pulling 0136fd5 on jmetz:convex_hull_image_qhull_precision_fix into 73bf701 on scikit-image:master.

[jni]

@jmetz @stefanv @ahojnnes I also can't get the cython files to compile automatically. I believe this was introduced with the "cythonize" syntax. What needs to be changed in our setup to get the files to be compiled with e.g. python setup.py install or make all?

[jni]

PEP8: image[nonzeros]

[jni]

@jmetz one final style comment, then it's in as far as I'm concerned!

[jmetz]

Thanks @jni and @JDWarner for the style comments.

[coveralls]

Coverage remained the same when pulling ef779d1 on jmetz:convex_hull_image_qhull_precision_fix into 73bf701 on scikit-image:master.

[jni]

@stefanv, I made a call to ignore the Travis error since it was obviously ancillary to the actual build. The previous build was fine, and this last commit only changed spacing between parentheses. Let me know if you'd rather I restart builds even in such cases.

[stefanv]

@jni Master looks happy, so I'm happy.