[SciPy-User] Difference in quality from different interpolation orders

[Zachary Pincus]

> Okay, thanks for that information. Interesting that higher-order interpolations could actually make the problem worse. I'd assumed that "higher order == more accurate" would hold true, but I guess it makes sense that for sharply discontinuous inputs, that breaks down. 

Higher-order has two relevant implications here:
 (1) More parameters to fit, which can either increase or decrease the plausibility of the interpolation (the classic overfitting vs. underfitting dilemma).
 (2) To have enough input to constrain the parameters, the fit is made over a larger window. Order-1 interpolation just looks at the two neighboring pixels to fit the interpolating line; higher orders look at more distant pixels, which can be helpful if the image is changing slowly, but in any case is slower as you've seen.

> I'm still very much inexperienced when it comes to scientific programming; I've only really done application programming and graphical work before. There's a lot of new background knowledge I have to get for many of these projects...

There's lots of people with good scientific imaging experience on this list. (My PhD's in image analysis for microscopy, for example, and I'm not the only one with similar experience.) Also check out the scikits.image project and list. Also, as far as I can tell, folks on the list are pretty receptive to general "how best to do task X" questions in addition to scipy-specific stuff.

Zach