Examples: contourApproach depends on envSampling modifying NDBC46022.h5

[ssolson]

example_contourApproach.py requires that example_envSampling.py has run and modified NDBC46022.h5.

This creates a required order to run the examples which is not clear to the intended audience (e.g. a new user). Further, while the user can specify the output H5 file name the duplication of the full buoy dataset does not seem ideal (e.g. the user now has 2 files of buoy data only differing by one having contour output from running example_envSampling.py). Finally, I believe this modification of the H5 file in place contributed to #29

I propose to address these issues and prevent future ones we:

1. Remove any dependency between examples
2. Restructure how we save contour data (e.g. I think we should look into having results separate from input)

[ssolson]

PR #36 established the data default locations expecting the user to run the examples from the examples folder. E.g. each example now has a relative path pointing to the repository example data.