Functions > Design of Experiments > Regression Analysis > General Multivariate Fit
General Multivariate Fit
multidfit(X, Y, G, F)—Returns the multivariate fit parameters for an arbitrary function F fitting the results recorded in matrix Y to the data found in matrix X.
The multidfit function acts as a command-line version of the Levenberg-Marquardt solver used in find and minerr, and as a multivariate version of genfit.
Arguments
X is an mxn design matrix or a matrix in which each column represents an independent variable. The number of columns n must be greater than one, and each column must have compatible units.
Y is a vector or a matrix of measured or simulated results with each row containing the results for each run or data point defined in X. When the rows do not all contain the same number of replicates, you must pad the empty elements of Y with NaNs. The elements of Y must have compatible units.
G is a vector of guess values, one for each parameter of vector u.
F(x, u) is a fitting function where x is a matrix of independent variable and u is a vector of parameters. For example:
F(x, u):= u0 + u1 * x01/2 + u2 * x12/3
To calculate the vector of parameters u, provide only the name of the fitting function F without any of its arguments to multidfit. Each variable in vector x must have units compatible with the corresponding column of X. The units returned by the fitting function F are compatible with the units of Y.