Re: [Scidavis-users] Peak labelling

[Joseph W. <jos...@mo...>]

Knut Franke wrote:
>> One thing that has be stuck is checking and displaying
>> peak maxima/minima.
> 
> I'm not sure I understand what you're trying to do; but in addition to what 
> Tilman already mentioned, there are column statistics (Analysis menu) which 
> give you the mathematical maximum/minum of a column; and multi-peak fitting 
> (when a graph is selected, this can be obtained from Analysis->Quick Fit) 
> which has the advantage of calculating peak positions based on all available 
> data, usually resulting in better treatment of discordant values.

In programs such as DPlot, you can click on an arbitrary part of a line
graph and the program creates a label with the co-ordinates of the graph
position. Dragging the label results in a line between the label and the
graph, which makes it easy to label particular "features", usually a
maximum or minimum values:

|    1008         2545
|     |            |
|     |           +++
|     +           + +
|    + +         +   +
|  ++  ++   +  ++     +   +
|+       +++ ++        +++
|__________________________
0     1000    2000     3000

(Usually, I'm only interested in the x-value, not the y-value.)

The problem with curve fitting it that it assumes a particular model,
whereas I'm aiming to label up the experimental data. If the model fits,
that is good, but it might not!
-- 
Joseph Wright