Custom Model Fitting

The modelfit2d function fits peak amplitudes across a series of 2D spectra to an arbitrary user-supplied equation. This is useful when none of the built-in models (exponential decay, recovery) describe your experiment.

The equation is specified as a string using standard mathematical syntax. The variable x represents the independent variable (e.g. a delay time, concentration, or frequency). All other symbols in the equation are treated as free parameters to be fitted.

<!– screenshot: assets/modelfit2d.png –>

Usage

using NMRAnalysis

# J-modulation experiment: amplitudes oscillate as A·sin(J·x)
modelfit2d(
    ["112", "113", "114", "115"],
    [0.1, 0.2, 0.3, 0.4],        # x values (e.g. delay times in seconds)
    "A*sin(J*x)",                 # model equation
    ["A" => 40.0, "J" => 0.5]    # parameter names and initial values
)

Parameters are specified as a vector of "name" => initial_value pairs. Good initial values improve convergence, especially for nonlinear models.

Common mathematical functions available in equations include sin, cos, exp, log, sqrt, and standard arithmetic operators.

Output

Clicking Save to folder writes all results to results.csv. The derived columns are named after the parameters given in the parameters argument (each with an _err uncertainty column). See Peak Lists and Output Files for the full format, and use summaryplot("output-folder/"; param=:paramname) to plot a parameter against residue number.