|
Packit |
67cb25 |
# -*- org -*-
|
|
Packit |
67cb25 |
#+CATEGORY: multifit
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
The following would also be nice additions to the multifit function suite
|
|
Packit |
67cb25 |
(see MS Excel regression output for example):
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
1. Produce the correlation coefficient (r) and other statistics.
|
|
Packit |
67cb25 |
2. Allow fit variable weighting (not observation weighting).
|
|
Packit |
67cb25 |
3. Allow for principal factor computations.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
The following features should be added to the nonlinear least squares routines:
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
1. Support for constraints
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2. robust nonlinear least squares and robust ridge regression
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
3. covariance matrices for ridge regression
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
4. min_workspace_p is dynamically allocated in gsl_multifit_linear_gcv_min - fix
|