Cls error masking

[nicholasturner1]

Addressing #66 : Consistently applying mask across error computation, cls error, and testing error/cls

Two further commits:

• Allowing the config.cfg file to be used 'out of the box' by changing Piriform fields to ISBI2012
• Requiring users to specify labels for datasets used during training. This was originally allowed for the full forward pass, but causes a vague error message for training sets, and the full forward pass no longer uses this section of the code.

[xiuliren]

@nicholasturner1 why do you change it to square loss? I think that auto is better for default setting.

[nicholasturner1]

The original thought was that it would be easier for new people to interpret/troubleshoot square error instead of having to take an extra step to figure out which error they're even using. Not a huge deal for me either way though. If you greatly prefer auto I'll swap it back

[xiuliren]

yes, I prefer auto. square loss corresponds to linear output. To help the new users, we should explain these in the comments in config.cfg file though.

[xiuliren]

@nicholasturner1 could you solve the conflict? current code did not pass the autotest in travis.

[nicholasturner1]

Nope, I hadn't changed things yet still (sorry...), but I also didn't merge it. @sachinravi14, do you know what the issues were?

[xiuliren]

@sachinravi14 did you make any fix of merging this pull request?

[sachinravi14]

No, I just merged and it worked for both synapse and semantic branches. Also, I think masking also needs to be fixed on test because I don't think the voxels are being counted correctly there.