PR: Implement support for "IPT" colourspace.

[MichaelMauderer]

Implements conversions between XYZ and IPT colourspaces, as well as an IPT hue correlate.

Closes #114.

[coveralls]

Coverage increased (+0.01%) when pulling ec38713 on feature/ipt into d44662b on develop.

[KelSolaar]

Overall it's nitpicking comments, the only real issue is for the IPT_to_XYZ definition name.

For consistency, I don't mind we do units tests also, I can write them if you want!

[KelSolaar]

Reminds me that I need to ask JetBrains for a PyCharm Open Source license!

[coveralls]

Coverage increased (+0.03%) when pulling 4eb03a4 on feature/ipt into d44662b on develop.

[MichaelMauderer]

I think I have addressed everything raised by you (good and valid points!)

I am not entirely sure about the referencing scheme that I now use.
Basically I am just citing the reference in the header (which is what it is there for). But it doesn't feel quite right. Maybe it should go into an extended description, where there is more detail and then the reference to the paper?

Oh neat Open Source licenses! Wasn't aware of that.

[KelSolaar]

Awesome! Thanks :) I will probably run PyCharm reformat on it later.

As for the extended description, I would suggest that you do that in the IPython Notebooks, for instance link the notebook like that: https://github.com/colour-science/colour/blob/feature/v0.3.3/colour/colorimetry/lightness.py#L19
and have fun: http://nbviewer.ipython.org/github/colour-science/colour-ipython/blob/master/notebooks/colorimetry/lightness.ipynb

I'm doing the luminance one as we are speaking, and it should close the colorimetry chapter.