DensityAnalysis and RMSF start/stop/step Issue Fix

Fixes #108

Changes:

Added self.n_frames to parallel.py
Updated DensityAnalysis conclude() to read self.n_frames instead of trajectory length
Updated RMSF to read self.n_frames in _conclude()
Added test_n_frames() to both DensityAnalysis and RMSF tests
Added test_accuracy() to test_rmsf.py to verify sub-selecting trajectories doesn't affect accuracy
Added new data files to test_density.py to use for testing time step sub-selections
Consolidated two tests in test_density.py into one test that verifies both the array values and the sum of their values for different trajectory lengths, blocks, and step sizes

PR Checklist