lp://qastaging/~nipy-developers/nipy/ffr
This experimental branch tests the feasibility of binding with numpy some libraries written in C for R or S-PLUS, such as the LME4 package (linear and nonlinear mixed models) by Douglas M. Bates. Our goal is typically to make it fast to run routines in a massively univariate fashion, by looping over voxels in C (via cython) rather than in R. This module needs R to be installed on the system to run as it will link dynamically with libR. This branch is distributed in GPL license due to the dependence on R, thus it is intended to stay aside from the Nipy mainline and be used as an optional plug-in.
- Get this branch:
- bzr branch lp://qastaging/~nipy-developers/nipy/ffr
Branch merges
Branch information
Recent revisions
- 50. By Alexis Roche <email address hidden>
-
Wrapped pyx module around a pure python interface to avoid weird crashes and for
legibility. - 49. By Alexis Roche <email address hidden>
-
Adding a macro to avoid calling FORTRAN routines via environment-
specific
names. - 48. By Alexis Roche <email address hidden>
-
The LME4 encapsulating routine now returns arrays for the fixed effects, their
variance matrices and the scale parameters. - 45. By Alexis Roche <email address hidden>
-
Computing the variance matrix of the fixed effect estimates in a mixed model.
Branch metadata
- Branch format:
- Branch format 6
- Repository format:
- Bazaar pack repository format 1 (needs bzr 0.92)