scipy - Constraints on fitting parameters with Python and ODRPACK -
i'm using odrpack library in python fit 1d data. works quite well, have 1 question: there possibility make constraints on fitting parameters? example if have model y = * x + b , physical reasons parameter can in range (-1, 1). i've found such constraints can done in original fortran implementation of odrpack95 library, can't find how in python.
of course, can implement functions such return big values, if fitting parameters out of bounds , chi squared big too, wonder if there right way that.
i'm afraid older fortran-77 version of odrpack wrapped scipy.odr
not incorporate constraints. odrpack95 later extension of original odrpack library predates scipy.odr
wrappers, , unclear legally include in scipy. there no explicit licensing information odrpack95, general acm toms non-commercial license.
Comments
Post a Comment