This should fix the problem of peak selection.
The problem was a combination of the highly pack diffraction data (900 points for a graph drawn on a comparable number of pixels).
This results, for a 5-pixels selection sensitivity, in the following selection left shift:
Additionally, the ~1 pixel window for selection was unreachable, as reported by Prof. Rohl.
In this fix, I kept both way of selecting a peak: from baseline or from top value. The reason is that in some few occasions, selecting by top proved to be a fast way to separate 2 very close peaks.
In any case, using the graph interface to change the x axis limits would work equally well.
BTW, I removed the unnecessary selection of ~zero background.
This should fix the problem of peak selection. The problem was a combination of the highly pack diffraction data (900 points for a graph drawn on a comparable number of pixels). This results, for a 5-pixels selection sensitivity, in the following selection left shift:
Additionally, the ~1 pixel window for selection was unreachable, as reported by Prof. Rohl.
In this fix, I kept both way of selecting a peak: from baseline or from top value. The reason is that in some few occasions, selecting by top proved to be a fast way to separate 2 very close peaks. In any case, using the graph interface to change the x axis limits would work equally well.
BTW, I removed the unnecessary selection of ~zero background.