[SciPy-User] spline interpolation and matplotlib interaction

Fabricio Silva silva@lma.cnrs-mrs...
Sat Jan 9 12:22:39 CST 2010

Hello folks,
has anyone ever thought of a piece of code that
- use a parametrization of a function with B-splines, i.e. with knots
and coefficients,
- let a user manipulate this parametrization through matplotlib (and
event handling mechanisms) ?

Within a scientific app, I would like to be able to handle simplified
representation of time-varying quantities that could be obtained from
measurement or from academic signals. I thought scipy.interpolation
module could help me in such a way :
- measurement of time signals
- using splrep to least-square fitting and identification of knots and
coefficients (with respect to a smoothing factor and a polynomial order)
- checking : show the approximated function with matplotlib and let the
user manually modify the control points.

It seems that I have not the sufficient understanding of b-splines as
the values output by splrep look strange even if the result of splev is

Fabrice Silva
Laboratory of Mechanics and Acoustics (CNRS, UPR 7051)

More information about the SciPy-User mailing list