--> atomsInstall("simplex")
This the simplex optimization method which performs optimization of non linear function. This optimization method doesn't use the derivative of the objective function. There is 2 versions of the simplex: - optim_nelder_mead: a classic script for optimization - step_nelder_mead: the call to the objective function is done outside of the script.
makes it available for scilab 5.4
Upload date : 2012-09-26 11:28:10 MD5 : 36dc17b6a24c1091b6fbee5ed8cfe32b SHA1 : f2d3809a263a3fd7ab1df3dfdcad54750653265c Downloads : 1256 File list
Linux 32-bit Automatically generated by the ATOMS compilation chain
Upload date : 2012-09-26 11:29:10 MD5 : 2ac454b601fdf964ef112c73bf5363d9 SHA1 : 7a497a4bae1a0f8e23852113040284c21b964e58 Downloads : 957 File list
Windows 32-bit Automatically generated by the ATOMS compilation chain
Upload date : 2012-09-26 11:29:00 MD5 : c48e2dd0f2cab21f6bbf93fd264bc5aa SHA1 : 6a831a93c0f617ec53b24cc2c08ec1a246be50da Downloads : 1530 File list
Windows 64-bit Automatically generated by the ATOMS compilation chain
Upload date : 2012-09-26 11:28:31 MD5 : fc55d820b88209a07fbcd0c95f6c3714 SHA1 : 004bfd041eaa179a58c16dbb3d555fc3c4da6417 Downloads : 2806 File list
MacOSX version Automatically generated by the ATOMS compilation chain
Upload date : 2012-09-26 11:31:20 MD5 : 5f74e5c7dda68e8e3efcb3b78285476d SHA1 : c4e41b6cfdfe03635095aac103906a021ef932cd Downloads : 1186 File list
Linux 64-bit Automatically generated by the ATOMS compilation chain
Upload date : 2012-09-26 11:30:11 MD5 : dce0bba9719b5d35a3bdba3aadeb3ab6 SHA1 : d4dff6e9b704e52caadefdff6794619306f7492c Downloads : 1624 File list
Hi, This is a message for users searching an derivative-free optimization function. Scilab 5 has a fminsearch function with Nelder-Mead algorithm, based on its "neldermead" module. The fminsearch function has full help pages and unit tests, with powerful options such as full exit status, full display control, an output function system (with a output function implementation), a plot function (with a default implementation), etc... This is why the "optim_nelder_mead" of the simplex module is very interesting for understanding the algorithm in detail (e.g. for academic purposes), but is unnecessary for industrial practice. The "step_nelder_mead" function on the other hand allows to evaluate the objective function outside Scilab, which is an interesting feature (although there are rare situations where this is required). Best regards, Michaël