www.gusucode.com > mbc 工具箱 matlab 源码程序 > mbc/@mbcmodel/@localresponse/RemoveOutliersForTest.m
function L = RemoveOutliersForTest(L,varargin) %REMOVEOUTLIERSFORTEST Removing outliers and refit response. % % REMOVEOUTLIERSFORTEST( LOCALRESPONSE, TESTNUMBER, LOCALSELECTION ) % REMOVEOUTLIERSFORTEST( LOCALRESPONSE, TESTNUMBER, LOCALSELECTION, DOUPDATE ) % % LOCALSELECTION can either be a set of indices, or a function name. % A outlier selection function must take the following form: % INDICES = MYMATLABFILE(MODEL, DATA, FACTORNAME); % The factors are the same as defined in DiagnosticStatistics. DATA % contains the factors as columns of a matrix FACTORNAME is a cell % array of the names for each factor. % If DOUPDATE is true then all response features are updated % % Example. For a local response LOCALRESPONSE % % remove first two data points and do not update response features % RemoveOutliersForTest(LOCALRESPONSE,1,1:2,false); % % find list of indices of removed data points % indices = OutliersForTest(LOCALRESPONSE,1); % % restore first data point % RestoreDataForTest(LOCALRESPONSE,1,1,false); % % restore all data % RestoreDataForTest(LOCALRESPONSE,1,':',false); % % update response features % UpdateResponseFeatures(LOCALRESPONSE); % % See also mbcmodel.localresponse.UpdateResponseFeatures, % mbcmodel.localresponse.RestoreDataForTest, % mbcmodel.localresponse.OutlierIndicesForTest % Copyright 2004-2010 The MathWorks, Inc. % Note it is also possible to use a special array to do standard types of % outlier removal (available from popmenus in model browser dialog). This % option is not documented because of its complexity RemoveOutliersForTest(L.Object, varargin{:});