XRambo is a tool for the extraction of quantitative parameters from a set of experimental data. It is not intended for general NMR data processing, but rather for the extraction of accurate quantitative information from the data. As experimentalists, we often wish to push our data to the limits. XRambo provides a means for extracting all of the information from a set of possibly suboptimal data, while at the same time providing realistic error estimates to prevent overzealous interpretation.
In some ways, the approach used by XRambo is similar to nonlinear curve fitting: we are given noise-corrupted data and a model function with adjustable parameters, and we must find what values of those parameters "fit" the data.
By adopting a Bayesian view of probability, we can go further and estimate not only the "best" or "most likely" values for the adjustable parameters, but also their joint distribution. This allows us to estimate uncertainties and correlations among the adjustable parameters in a very direct way.
XRambo uses the Metropolis Monte Carlo algorithm [link edited: LCM] to directly sample points in the parameter space of interest to the user. This generates a cloud of points which form an approximation to the joint probability density function of the parameters given the data.
Although XRambo is heavily optimized for fitting linear combinations of exponentially-decaying sinusoids, it does provide the option of specifying arbitrary user-defined model functions.
XRambo has been implemented in an X/Motif interface.