modhpsol.spad line 1 [edit on github]
This package solves (multipoint) Hermite-Pade interpolation problem over $Z_p
$, where $p
$ is machine sized prime.We use quadratic algorithm from 94 Beckermann and Labahn paper (with modification to handle multipoint case).
naiveBeckermannLabahn(f, degs, sigma, p)
solves modulo $p
$ Hermite-Pade interpolation problem of order $sigma
$. $degs
$ specifies degree bounds for solution. The first element of result contains sigma
-basis. The second gives defects of elements of sigma
-basis. The third one contains residuals --
we stop updating solution if this would violate degree bounds, so residual is zero for elements of sigma
-basis which are within bounds, but non-zero otherwise.
naiveBeckermannLabahn(f, degs, sigma, p)
solves modulo $p
$ Hermite-Pade interpolation problem of order $sigma
$. $degs
$ specifies degree bounds for solution. The first element of result contains sigma
-basis. The second gives defects of elements of sigma
-basis. The third one contains residuals --
we stop updating solution if this would violate degree bounds, so residual is zero for elements of sigma
-basis which are within bounds, but non-zero otherwise.
main solver routine
naiveBeckermannLabahnMultipoint(f, degs, sigma, p)
solves modulo $p
$ multipoint Hermite-Pade interpolation problem of order $sigma
$. $degs
$ specifies degree bounds for solution. The first element of result contains sigma
-basis. The second gives defects of elements of sigma
-basis. The third one contains residuals --
we stop updating solution if this would violate degree bounds, so residual is zero for elements of sigma
-basis which are within bounds, but non-zero otherwise.
naiveBeckermannLabahnMultipoint(f, degs, sigma, p)
solves modulo $p
$ multipoint Hermite-Pade interpolation problem of order $sigma
$. $degs
$ specifies degree bounds for solution. The first element of result contains sigma
-basis. The second gives defects of elements of sigma
-basis. The third one contains residuals --
we stop updating solution if this would violate degree bounds, so residual is zero for elements of sigma
-basis which are within bounds, but non-zero otherwise.
reduceBasis skips
elements of sigma basis which does not satisfy degree bounds and puts sigma-basis into canonical form.