// Scilab ( http://www.scilab.org/ ) - This file is part of Scilab // Copyright (C) 200 - INRIA - Carlos Klimann // // This file must be used under the terms of the CeCILL. // This source file is licensed as described in the file COPYING, which // you should have received as part of this distribution. The terms // are also available at // http://www.cecill.info/licences/Licence_CeCILL_V2.1-en.txt // function [s,index] = nanmax(x,orient) // //This function gives for a real or a numerical matrix x a his largest //element m (but ignoring the NANs). // //For x, a numerical vector or matrix, m=nanmax(x) returns in scalar m //the largest element of x (ignoring the NANs). The form [m,index] = //nanmax(x,orient) gives in addition of the value of the largest element //of x (ignoring the NANs) in scalar m, and the index of this element //in x, as a 2-vector. // //m=nanmax(x,'r') gives in the 1xsize(x,2) matrix m the largest elements //(ignoring the NANs) of each column of x. If the form //[m,index]=nanmax(x,'r') is used, the elements of the 1xsize(x,2) //matrix index are the indexes of the largest elements (ignoring the //NANs) of each column of x in the corresponding column. // //m=nanmax(x,'c') gives in the size(x,2)x1 matrix m the largest elements //(ignoring the NANs) of each row of x. If the form //[m,index]=nanmax(x,'c') is used, the elements of the size(x,2)x1 //matrix index are the indexes of the largest elements (ignoring the //NANs) of each row of x in the corresponding row. // // // correction of bug 1239 (Bruno) [lhs, rhs] = argn(0) if rhs == 0 then error(msprintf(gettext("%s: Wrong number of input arguments: %d to %d expected.\n"),"nanmax",1,2)) end if rhs==1 then orient="*",end if orient==1 then orient="r",end if orient==2 then orient="c",end if x==[]|(size(x,"*")==1&isnan(x)) then s=[],index=[],return,end isn=isnan(x) x(isn)=-%inf [s,index]=max(x,orient) s(find(and(isn,orient)))=%nan endfunction