A maintenance PR to write a few portions of barycentric precompute optimization strategy differently, along with a few cleanups and then cross testing it with a Lagrange Interpolated value from Go-IPA.
More elaborately
Avoiding typecasting and directly using fromInt instead converting too many times using uint64 and BigInts, this was also an issue for the CI check earlier.
Fixing an index issue in PrecomputedWeights, the length of invertedDomain should be 2 less than that of barycentric weights, because there will be 2 such instances of 1/k and -1/k where k will essentially be 0, so we need to avoid those indexes in order to compute correctly.
Fixing some Endianness issues, as the Verkle Spec follows only littleEndian when it comes to ser/de of Scalars
Just some cosmetic things that came up now and then.
A maintenance PR to write a few portions of barycentric precompute optimization strategy differently, along with a few cleanups and then cross testing it with a Lagrange Interpolated value from Go-IPA.
More elaborately
fromIntinstead converting too many times using uint64 and BigInts, this was also an issue for the CI check earlier.littleEndianwhen it comes to ser/de of Scalars