Project page: http://sourceforge.net/projects/pythonlabtools/
The main large modules in analysis are the fitting_toolkit.py and C2Functions.py.
fitting_toolkit.py provides a very strong basis of support for generalized non-linear least-squares fitting, including bootstrapping (resampling), weighted fits, correlated data fits, inverse Hessian, Levenberg-Marquardt, and singular-value-decomposition Hessian searches.
C2Functions.py provides a framework for carrying out numerical operations on smooth (twice-differentiable) functions, especially (but not limited to) cubic splines. It provides fast root finding (using inverse quadratic extrapolation) and fast adaptive integration (10th order convergence, if the derivatives are well behaved).