Fix: precision in hazard curve and UHS output

[chrisdicaprio]

The precision for the y-axis of hazard curves and UHS curves in the exported csv file is fixed. However, for low probability (hazard curves) or low acceleration (UHS) the provided precision may not be enought to differentiate one point from the next.

For example. Auckland, vs30=400m/s, 86% in 50

Hazard Curve

UHS

Propose using scientific notation with a fixed minimum number of sig figs (see tooltip formatting)