Fix boundaries for first polar pixel above z=2/3

[smcguire-cmu]

For checking whether to use polar or eq coordinates, we check if eq_z > 2/3 which isn't entirely accurate when looking at neighbor pixels. There are possible cases where this is true, but the polar values of k and kp are invalid. We previously cut off these values, but this causes some missing values in the first polar pixel above z=2/3 that borders the phi=0 line, as demonstrated in #2.

This updates the polar function to use the equatorial function when the values of k and kp are invalid for generating polar boundaries, which fixes the problem. This however does also cause overlap of edges in the pixels with 7 neighbors, so we may want to investigate changing the boundary values for those cases.

fixes #2

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[hombit]

Looks great