The Peaks dialog

This is what the Analyze Peaks... dialog looks like:

Peaks and troughs are found by a simple threshold mechanism. A peak is recognized as an increase of at least the minimum peak height within at most the maximum peak half width, followed by a decrease of at least the minimum peak height within at most the maximum peak half-width. Troughs are similarly recognized. These thresholds are set by the sliders in the top panels. The t scale and V scale arrows allow you to change the range of the sliders.

Below the thresholds panel is a marks panel. It tells xdatplot where to search for peaks, and how to mark them. If there are no marks in your file and the region is not defined, the entire file is searched. If the region is defined, only it is searched. If there are 0 and 1 marks, the search starts at the first 0, then searches up till the next 1, skips until the next 0, searches until the next 1, etc. The "Begin at Mark" and "End at Mark" fields can be changed from 0 and 1 to something else. This is useful for blocking out noisy regions of your data.

The results of the peaks search are reported by setting marks in your data. "Mark Peaks with" and "Mark Troughs with" control how this is done. If the mark peaks field is blank, peaks aren't marked. If it contains one character, the peaks are marked with that character. If it contains two characters, the left and right edges of the peak are marked with those characters. If it contains three characters, the left edge, peak, and right edge are marked.

A comment is attached to each mark indicating the t and V values there. The comments of marks at the center of peaks or troughs also give the height and area of the peak. (Note: the height is not the same as the data value at the peak; it is the difference between that value and the average of the data values at the left and right bounds.) The exact format of these comments is controlled by the markCommentFormat3 resource.

Leon Avery (leon@eatworms.swmed.edu)