This is not intended to be a completed transition since it leaves the
libraries unusable to external client code, but rather as a step towards
use of math/rand/v2. This initial step allows repair of sequence change
failures without having to worry about API difference.
The links have been redirecting to `pkg.go.dev` for a while now. In the
case of the main README, that made the badge a virtual duplicate of the
"go.dev reference" one.
Changes made in dsp/fourier/internal/fftpack break the formatting used
there, so these are reverted. There will be complaints in CI.
[git-generate]
gofmt -w .
go generate gonum.org/v1/gonum/blas
go generate gonum.org/v1/gonum/blas/gonum
go generate gonum.org/v1/gonum/unit
go generate gonum.org/v1/gonum/unit/constant
go generate gonum.org/v1/gonum/graph/formats/dot
go generate gonum.org/v1/gonum/graph/formats/rdf
go generate gonum.org/v1/gonum/stat/card
git checkout -- dsp/fourier/internal/fftpack
* optimize: Change initialization, remove Needser, and update Problem function calls
We need a better way to express the Hessian function call so that sparse Hessians can be provided. This change updates the Problem function definitions to allow an arbitrary Symmetric matrix. With this change, we need to change how Location is used, so that we do not allocate a SymDense. Once this location is changed, we no longer need Needser to allocate the appropriate memory, and can shift that to initialization, further simplifying the interfaces.
A 'fake' Problem is passed to Method to continue to make it impossible for the Method to call the functions directly.
Fixes#727, #593.
* optimize: Refactor gradient convergence and remove DefaultSettings
The current API design makes it easy to make a mistake in not using the DefaultSettings. This change makes the zero value of Settings do the 'right thing'. The remaining setting that is used by the DefaultSettings is to change the behavior of the GradientTolerance. This was necessary because gradient-based Local methods (BFGS, LBFGS, CG, etc.) typically _define_ convergence by the value of the gradient, while Global methods (CMAES, GuessAndCheck) are defined by _not_ converging when the gradient is small. The problem is to have two completely different default behaviors without knowing the Method. The solution is to treat a very small value of the gradient as a method-based convergence, in the same way that a small spread of data is a convergence of CMAES. Thus, the default behavior, from the perspective of Settings, is never to converge based on the gradient, but all of the Local methods will converge when a value close to the minimum is found. This default value is set to a very small value, such that users should not want a smaller value. A user can thus still set a (more reasonable) convergence value through settings.
Fixes 677.
