Fast Fitting with GPU acceleration

pawel.wzietek

This project, based on an older version of the ''GenCurvefit'' package by Andrew Nelson (https://www.wavemetrics.com/project/gencurvefit), uses differential evolution algorithm for curve fitting but adds a possibility of GPU acceleration (OpenCL interface via IgorCL XOP, https://www.wavemetrics.com/project/IgorCL) for extremely fast fitting of complex models. The demo projects included can be used as a starting point to develop your own models.

Project Details

Current Project Release

Release File: GenOpCL_102_0.zip (393.17 KB)
Version: 102
Version Date:
Version Major: 1
Version Patch Level: 02
Release Notes:

Minor update in the demo experiments (other files not modified) to make them work correctly on AMD hardware: apparently the AMD compiler has issues with code exported from notebooks because the line breaks are coded with "\r" instead of "\r\n". I added, in the function getCLcode, the following line:
clcode=replacestring("\r", clcode, "\r\n")
This should make all compilers happy.

View All Releases
Analysis Curve Fitting

Forum

Support

Gallery

Igor Pro 10

Learn More

Igor XOP Toolkit

Learn More

Igor NIDAQ Tools MX

Learn More

AltStyle によって変換されたページ (->オリジナル) /