Skip to content
This repository


Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP


Fixed README.rst, and pValue deleted
Maxfev increased to maxfev = 250*(len(params0)+1)


A new input method implemented using argparge (see REAME.rst)
Corrected a bug in the calculation of the Jacobian. Now it is possible
to calculate the fit using two or more functions and data even with
analytical derivatives.

Corrected a bug in the analytical derivatives for data.X = 0,
and derivative as a constant:
- derivative = "%s*%s/%s" % (derivative, var, var)
+ derivative = "%s*log(exp(%s))" % (derivative, var)


Added option -h --held, to keep a parameter fixed to a value, i.e. -h…
… q=.3


Fix a number of bugs on the use of the options (-s and -d)
Introduced dFunc as an alias for analytical derivatives
Legend eliminated (to be improved)
Added a new separator "_and_" for multiple functions and filenames (to be documented yet)


Fit multiple datasets with different functions, and
at least one common fitting parameter.

Example of use with 2 gaussian and the same sigma:
bestFit -f data1.dat,data2.dat -t "a*exp(-((x-x01)/sigma)**2)","b*exp(-((x-x02)/sigma)**2)" -p a,b,x01,x02,sigma -i 1,1,1,0,1.

To be added to README.rst

Use Qt4 backend for Matplotlib if available, as it enables to
change all the figure elements, such as title, axis, colors etc.
At the moment there is a bug in it (matplotlib 1.0.1).


Added blank lines in README.rst


Command 'exec' eliminated using
f = sympy.simplify(function_string) and


Deleted toUniformLogDistribution from the repo


A few improvements:
* Added scipy.special functions (for instance, bessel, gamma, etc)
for symbolic calculation of derivatives and in the function definition.
* Solved a problem with function names, as in f = gamma(x) + polygamma(0,x).
* Added a test function in getAnalyticalDerivatives
* Added a check on NumExpr which works only on limited functions
* Corrected a few bugs on use of the 3rd column as a errors of data
Something went wrong with that request. Please try again.