LMA on Python

Python, actually Scipy, provides the Levenberg-Marquardt algortihm for solving least-square problem. It's called leastq in scipy.optimize package. There are a lot's optimization algorithms.

http://docs.scipy.org/doc/scipy/reference/optimize.html

Basically, It was coded by fortran. http://www.netlib.org/minpack/

You can download the test code for fortran version form http://www1.fpl.fs.fed.us/minpack.ls.test.html

You can use it directly by calling "scipy.optimize.leastsq" after including "from scipy.optimize import leastsq"

These are good simple examples.

http://xinqingpeng.blogspot.com/2007/07/scipy-fitting-2-d-data.html

http://www.tau.ac.il/~kineret/amit/scipy_tutorial/

http://www.scipy.org/scipy_Example_List#head-4c436ae0085d9a56056425d11abff4ccdd3d3620

Explanation in Chinese http://qzone.qq.com/blog/55287019-1232730815

Reminder:

Be careful the arrary usage of input data x and y.