www.gusucode.com > mbcexpr 工具箱 matlab 源码程序 > mbcexpr/@cgnormfunction/fit_table.m
function T = fit_table(LU,X,Y,M); %FIT_TABLE Returns data for comparison plot in breakpoint editor. % % T = FIT_TABLE(T,X,Y,M) % X & M are arrays of the same size. Y is ignored and exists only to % provide a uniform interface with cglookuptwo objects. % M is the value of the model at points X. % T is an array of the same size as the values of the normaliser for this % lookup table, and is the best fit for the model at these points. % Should be used after return_data has run. The reason for splitting them up % is that we need to find new table values each time we edit the table, whereas % we need to evaluate the model only when the input variables change. % Copyright 2000-2004 The MathWorks, Inc. and Ford Global Technologies, Inc. cgm = cgmathsobject; xNormaliser = LU.Xexpr; Xinput = xNormaliser.get('x'); BP = xNormaliser.get('breakpoints'); V = xNormaliser.get('values'); BP = eval(cgm,'linear1',V(:),BP(:),[0:V(end)]'); BP = mbcmonotonic(BP); A = eval(cgm,'values_regression1',X(:),M(:),BP); T = eval(cgm,'linear1',BP,A,X(:));