Peak Separator

Separate Merged Peaks Using Peak Models

The Peak Separator code uses a Markov Chain Monte Carlo search which, using idealised peak shapes, attempts to deconvolve overlapped peak regions into their separate constituent peaks.

This routine is also suitable for accurately fitting model shapes to single peaks in order to calculate precise intensities.

Options Peak Separator Parameters Min. Number of peaks is by default set to one, it is not possible to set this to a value less than one. Max. Number of peaks is by default set to one, increasing this value allows the search routine to fit more models. The best fit may be found with fewer than the maximum number models. Higher numbers slow the routine, and setting this value to 0 allows the routine to (effectively) fit unlimited peaks. Only pick positive peaks. If you are not interested in negative peaks, removing the possibility of fitting negative peaks can reduce search time. Peak Model fits the spectra with either a Gaussian peak model or a Lorentzian peak model.

Options Region Peak List choose which peak list newly picked peaks should be added to. Peaks picked using this method will have their details appended with ‘PeakSepartor’ so you know where they came from. Region Table shows which area of the current spectrum is about to be searched. Add Region. Once an area of spectra has been highlighted clicking this button will pass it’s details on to the Peak Separator. Reset All will reset all search parameters. Separate Peaks will run the Peak Separator code with your current settings. This may take a few minutes to run, depending on the size of the spectral region being searched, the number of peaks being fitted and the speed of your machine. Please wait while this completes.

After a successful Peak Separation run, the found peaks will be added to the selected peak list. These peaks intensties (volume) have been found using the peak model selected.

Advanced Settings Tab Rate affects the speed of the Markov Chain Monte Carlo routine. A smaller value results in longer execution, but possibly higher quality results. The default setting is deemed sensible for the majority of runs. Line Width offers a finer degree of control over maximum and minimum peak widths for each dimension. The default values are very stupid and could do with re-checking for each experiment. Re-Pick Entire Peak List if you would like to use the Peak Separator to repick every peak in your peak list, try this option - but note that this may take a very long time!

Main Panel

button Clone: Clone popup window

button Help: Show popup help document

button Close: Close popup

Peak Separator

Documentation missing

Peak Separator Parameters

int Min. number of peaks: Minimum number of peaks to find (must be > 0)

int Max. number of peaks: Maximum number of peaks to find (0 is unlimited - not recommended)

radio False: Search for both positive and negative intensity peaks

radio True: Limit search to only positive peaks

radio Gaussian: Choose a Gaussian model peak shape to fit to peaks

radio Lorentzian: Choose a Lorentzian model peak shape to fit to peaks

Region

Region that search will limit itself to

pulldown Peak List: Select which peak list new peaks are to be added to

Table 1
dim. Documentation missing
start (ppm) Documentation missing
end (ppm) Documentation missing
actual size Documentation missing

button Add Region: Add selected specrtral region

button Separate Peaks: Run peak search now

Advanced Settings

Documentation missing

Rate:

float Rate of MCMC step size change: Rate effects speed of run, smaller values take longer but may produce better results

Line Width (Hz):

(optional - repick entire peak list)

pulldown *None*: Select which peak list to repick (new peaks will be put into a new peak list)

button Repick Peak List: Repick selected peak list into a new peak list.

button Separate Peaks: Run peak search now

Table Of Contents

Previous topic

Peak : Selected Peaks

Next topic

Peak : Peak Finding

This Page