Sidebar widget for peak rejection

See #482 (WIP)

This MR adds a new sidebar widgetd, SubframeReject, in which the user can use both heuristics and machine learning algorithms to reject peaks on based on peak statistics. So far, the heuristics part has been implemented. Currently, the user can plot a histogram of intensity, sigma or strength, select a window of the resulting histogram, and filter the peaks within that window so that they can be found easily. Peaks can then be manually flagged as outliers.

Statistical outlier rejection has yet to be implemented, either by some Bayesian method, or some outlier detection method.

Resolves #138 (closed)