FORTRAN SOFTWARE CATALOG

Repository owner: Federico E. Benelli (fedebenelli)

fortimize

A common interface to optimization algorithms including Nelder-Mead and gradient descent.

Repository · Pinned source (3431b656cd86) · Try in playground · Checked package example

GAMS classifications

About this package

Category
Roots and optimization
License
Not specified upstream (manifest placeholder)

Topics: derivative-free, fortran, fortran-, gradient-descent, nelder-mead, optimization, optimization-algorithms

Package dependencies

Resolved GitHub repositories used by the verified playground build, combining direct and indirect dependencies. This list does not describe internal module dependencies or native system libraries.

Required by this build

Build record verified 2026-10-27.

Used by (0 selected packages)

No other selected build records this repository as a dependency. This is not a measure of usage outside the catalog.

Reviewed procedures

These mappings have pinned source evidence. A checked package example may exercise only some of them; source review is not a numerical benchmark.

minimize with NelderMead

Module: ForTimize

G1b

G1b: verified

Derivatives / callbacks: none

The existing catalog quadratic program is an independently written API example. Upstream licensing is unspecified; no upstream example code is copied here. Dedicated G1b consumer checks twelve derivative-free solves: coupled positive-definite quadratics in 2/3/5 variables and the two-variable Rosenbrock problem, each from three starts. Known unique minimizers and zero objective supply independent acceptance criteria, and the returned objective is recomputed. Uses the actual kcount evaluation limit; optimizer termination status is not exposed by this wrapper. No upstream example code is copied. Other methods, nonsmooth/global optimization, scaling extremes and concurrent calls are not verified.

Pinned source: src/minimize.f90 · src/gradient_free/nelder_mead.f90

Author examples/tests: example/gradient.f90

Load checked GAMS driver: gams-fortimize-nelder-reference

minimize with GradientDescent

Module: ForTimize

G1b

G1b: verified

Derivatives / callbacks: user gradient

Fixed-iteration gradient descent with a user gradient. The dedicated independent driver checks 240 SPD-quadratic trajectories against spectral powers and exact callback counts. At this pin the returned objective belongs to the iterate before the final position update; the driver explicitly verifies this and recomputes the endpoint objective. convergence_tolerance is not used by this fixed-count implementation. No adaptive stopping or nonquadratic convergence claim; upstream licensing remains unspecified.

Pinned source: src/minimize.f90 · src/with_derivatives/gradient_descent.f90

Author examples/tests: example/gradient.f90

Load checked GAMS driver: gams-fortimize-gradient-reference