Adds in peak-find-mode option with slow, average, and fast options. This controls how peaks are split for fitting data and only really has an effect on data with many peaks. Slow should be used if only a few peaks are ever being fit, but for cases where the data is very complex and speed is desired, fast can be supplied to split the data up to fit subsections of the data. The downside of this is certain overlapping peaks may be incorrectly split and the fits are suboptiomal.