symfit 0.3.0

symfit; data fitting for the practical man.

Existing fitting modules are not very pythonic in their API and can be difficult for humans to use. This project aims to marry the power of scipy.optimize with the readability of SymPy to create a highly readable and easy to use fitting package which works for projects of any scale.

symfit makes it extremely easy to provide guesses for your parameters and to bound them to a certain range:

a = Parameter(1.0, min=0.0, max=5.0)

To define models to fit to:

x = Variable()
A = Parameter()
sig = Parameter(1.0, min=0.0, max=5.0)
x0 = Parameter(1.0, min=0.0)

# Gaussian distribution
model = A * exp(-(x - x0)**2/(2 * sig**2))

To execute the fit:

fit = Fit(model, xdata, ydata)
fit_result = fit.execute()

And finally, to evaluate the model using the best fit parameters:

y = model(x=xdata, **fit_result.params)

For more examples, check out the docs at http://symfit.readthedocs.org/.

New in 0.3.0:

You can now name dependent variables:

x, y = variables('x, y')
model = {y: a * x**2}

Allowing for very clear assignment of data and standard deviations:

fit = Fit(model, x=xdata, y=ydata, sigma_y=sigma)
fit.execute()