DensityAnalysis Normalization Fix

[nawtrey]

Fixes #108

Changes made in this Pull Request:

• Correct n_frames assignment in init()
• Add test function to test_density.py to verify correction

PR Checklist

• [ ] Tests?
• [ ] Docs?
• [ ] CHANGELOG updated?
• [ ] Issue raised/referenced?

[nawtrey]

@orbeckst Do you know of another test trajectory with more frames than the current one in test_density.py (GRO_MEMPROT, XTC_MEMPROT)? There are only so many ways you can slice 5 frames.

[orbeckst]

For local testing: install MDAnalysisData

• https://www.mdanalysis.org/MDAnalysisData/ifabp_water.html
• https://www.mdanalysis.org/MDAnalysisData/yiip_equilibrium.html

For testing (available via MDAnalysisTests.datafiles)

• resname TIP3: "waterPSF","waterDCD" (<DCDReader ~/anaconda3/envs/mda3/lib/python3.6/site-packages/MDAnalysisTests/data/watdyn.dcd with 10 frames of 15 atoms>)
• resname WAT: PRMncdf, NCDF: <NCDFReader ~/anaconda3/envs/mda3/lib/python3.6/site-packages/MDAnalysisTests/data/Amber/bala.ncdf with 30 frames of 2661 atoms>

EDIT: Possibly more in MDAnalysisTests (investigate):

• PSF_TRICLINIC, DCD_TRICLINIC
• PDB_xvf, XTC_xvf
• PRM_NCBOX, TRJ_NCBOX