Updating prebuilt_nuc_data.h5 with mass.mas16 data

[aruzzo]

Addresses issue #18:

• Updating prebuilt_nuc_data.h5 to incorporate new atomic mass data from mass.mas16

I'm not sure what the verification process is for a big change to fundamental data, but I updated the few remaining failing tests from the original mass update PR (failing tests here) and ran travis-run-tests.sh and it passes the check (CI Test).

The nuc data file was updated by editing the existing prebuilt_nuc_data.h5 file with pyne functions. I edited the make_atomic_mass function in pyne/dbgen/atomic_mass.py to pass instead of return if the atomic_mass node already exists (which it does). Then, it deletes the prebuilt atomic mass node and reads in the data from mass.mas16. I then similarly edited pyne/dbgen/materials_library.py to pass instead of return, delete the prebuilt nodes, and then write in the new material library data.

[pshriwise]

I'm not sure how long it's been since this file was last changed, but as a start it'd be nice to have some kind of human-readable summary here in the PR with the changes to relevant datasets in the file as some kind of record for the scope of the changes and exact value updates if they need to be referenced in the future. A tool like h5diff might be able to help with that.

[aruzzo]

I used h5diff to try to compare the different nuc data files but the atomic_mass group now has a different size, so it's unable to compare them:

dataset: </atomic_mass> and </atomic_mass> Not comparable: </atomic_mass> has rank 1, dimensions [3466], max dimensions [18446744073709551615] and </atomic_mass> has rank 1, dimensions [3549], max dimensions [18446744073709551615]

[aruzzo]

I made a gist (atomic_mass gist) that highlights the differences in atomic mass between the old and new nuc_data.h5 files. It also prints the new nuc_ids that are present in the new file but not in the old one. For material_library, h5diff works in showing the difference for them. I don't know if there's a way to format the output, so I just piped it to a text file (mat_lib gist).

[gonuke]

Thanks @aruzzo - we are finally taking care of this during our hack-a-thon.