Feat normal-score-transform for DSI

[rhugman]

Allows for users to specify normal-score transformation and/or log-transformation of observation data when calling pyemu.EnDS.prep_for_dsi().

Notes:

• included python native 1D interpolation Class in helpers with the sole purpose of avoiding from scipy.interpolate import interp1d. I stress tested against scipy's interp1d. Differences start to creep in when extrapolating far from the original x-values. It may be worth using scipy instead (or someone smarter than me taking a crack at it...).

[coveralls]

coverage: 79.309% (-0.01%) from 79.322%
when pulling 8bece65 on rhugman:feat_normtransf
into ca8cb77 on pypest:develop.