Give me little more guide before making PR

[sshojiro]

I have attempted to make a PR https://github.com/CCampJr/pyMCR/pull/5. However, as you can see, codecov shows some errors. Could you give me little instruction to fix this?

Sorry to disturb you.

[CCampJr]

@sshojiro

Thank you very much for your hard work!

• Make the PR to the 0.3.X branch since that's what you forked from
• In general, I try to avoid using sklearn as this adds a requirement for it to be installed for future users. PCA, for example, can be calculated using SVD from numpy.linalg.svd
• In the demo, for example, avoid using words like "outstanding" or "great" -- we want to be as objective as possible.
• Please move the two metrics into a new python file called rank.py as these are effectively rank-finding methods. I intend for metric.py to only contain error function metrics that can be used by MCR for minimization.

Thank you very, very much for your contribution. I look forward to downloading your code and testing it out.

[sshojiro]

@CCampJr I also appreciate your great contribution so much. I've been thinking to start some research projects involving MCR-ALS. So, this repository will be a great help.

Thank you for your comments. After fixing your feedbacks, I'll make a PR again.

Well, as for the second comment,

In general, I try to avoid using sklearn as this adds a requirement for it to be installed for future users. PCA, for example, can be calculated using SVD from numpy.linalg.svd

I think you should add this description to CONTRIBUTING.rst. Hope that you will do so when you have time. Thank you in advance.

[CCampJr]

@sshojiro

Good point: I will add a note to avoid external libraries unless absolutely necessary to the contributing guide-- thanks!

[CCampJr]

Just added a statement to CONTRIBUTING.rst

[CCampJr]

Thanks for the PR-- closing this issue