SDEXPORT Export data set to file or pipeline for out-of-Matlab execution Export data set to a comma-separated file: SDEXPORT(DATA,FILENAME) SDEXPORT(+DATA,FILENAME) % export a data matrix only, no header Export trained pipeline for execution out of Matlab: SDEXPORT(P,FILENAME,options) INPUT DATA SDDATA object P Pipeline object FILENAME Filename of the pipeline file OPTIONS Exporting data sets 'format' Export format (default: 'csv') 'csv' Comma-separated text 'arff' Weka ARFF format 'c45' C45 compatible data set 'test' When exporting in C45 format, save as a test set (.test extension) 'columns',COL Specify properties saved in the csv file (default: {'data','lab'}) 'no header' Do not write a header to .csv file 'numformat',F Numercial format specifier (for sprintf) (def: '%f') Exporting pipelines 'header' Save pipeline as a C-header file to be included in C source 'cubert' Export Cubert Utils plugin (xml file) DESCRIPTION SDEXPORT writes data set DATA into comma-separated file. Only data matrix and labels are stored. Use SDIMPORT to load SDDATA from text file. See file header for the exact SDIMPORT command needed to load the data. For text-based formats, 'numformat' option allow to adjust precision. For example, we may use 'numformat','%e' to use exponent notation. SDEXPORT may save the trained pipeline P into a binary file which may be executed by the perclass.dll runtime linked to a custom application. With 'header' option, SDEXPORT saves the pipeline as a C header file which may be directly included in C/C++ project. It is also possible to export Cubert Utils plugin (.xml file) which can be directly loaded into Cubert Touch environment. SEE ALSO SDIMPORT READ MORE http://perclass.com/doc/guide/deployment.html
sdexport
is referenced in examples: