Refactor in preparation to resume using the stereographic library for Newton's method.

3 files changed, 48 insertions, 40 deletions

diff --git a/langevin.cpp b/langevin.cpp
index 1a0e104..e4b6c8d 100644
--- a/langevin.cpp
+++ b/langevin.cpp
@@ -4,22 +4,11 @@
 #include <getopt.h>
 #include <random>
 
-#include <eigen3/Eigen/Core>
-#include <eigen3/unsupported/Eigen/CXX11/Tensor>
-
 #include "pcg-cpp/include/pcg_random.hpp"
 #include "randutils/randutils.hpp"
 
 #include "complex_normal.hpp"
-#include "tensor.hpp"
-
-#define PSPIN_P 3
-const unsigned p = PSPIN_P; // polynomial degree of Hamiltonian
-
-using Scalar = std::complex<double>;
-using Vector = Eigen::VectorXcd;
-using Matrix = Eigen::MatrixXcd;
-using Tensor = Eigen::Tensor<Scalar, PSPIN_P>;
+#include "p-spin.hpp"
 
 using Rng = randutils::random_generator<pcg32>;
 
@@ -36,32 +25,6 @@ Vector initializeVector(unsigned N, double a, Rng& r) {
   return z;
 }
 
-std::tuple<Scalar, Vector, Matrix> hamGradHess(const Tensor& J, const Vector& z) {
-  Matrix Jz = contractDown(J, z); // Contracts J into p - 2 copies of z.
-  Vector Jzz = Jz * z;
-
-  double f = factorial(p);
-
-  Matrix hessian = ((p - 1) * p / f) * Jz;
-  Vector gradient = (p / f) * Jzz;
-  Scalar hamiltonian = (1 / f) * Jzz.dot(z);
-
-  return {hamiltonian, gradient, hessian};
-}
-
-std::tuple<double, Vector> WdW(const Tensor& J, const Vector& z) {
-  Vector gradient;
-  Matrix hessian;
-  std::tie(std::ignore, gradient, hessian) = hamGradHess(J, z);
-
-  Vector projectedGradient = gradient - (gradient.dot(z) / (double)z.size()) * z;
-
-  double W = projectedGradient.cwiseAbs2().sum();
-  Vector dW = hessian.conjugate() * projectedGradient;
-
-  return {W, dW};
-}
-
 Vector langevin(const Tensor& J, const Vector& z0, double T, double γ0,
                 std::function<bool(double, unsigned)> quit, Rng& r) {
   Vector z = z0;
diff --git a/p-spin.hpp b/p-spin.hpp
new file mode 100644
index 0000000..1ed4d8e
--- /dev/null
+++ b/p-spin.hpp
@@ -0,0 +1,43 @@
+#pragma once
+
+#include <eigen3/Eigen/Core>
+#include <eigen3/unsupported/Eigen/CXX11/Tensor>
+
+#include "pcg-cpp/include/pcg_random.hpp"
+#include "randutils/randutils.hpp"
+
+#include "tensor.hpp"
+
+#define PSPIN_P 3
+const unsigned p = PSPIN_P; // polynomial degree of Hamiltonian
+
+using Scalar = std::complex<double>;
+using Vector = Eigen::VectorXcd;
+using Matrix = Eigen::MatrixXcd;
+using Tensor = Eigen::Tensor<Scalar, PSPIN_P>;
+
+std::tuple<Scalar, Vector, Matrix> hamGradHess(const Tensor& J, const Vector& z) {
+  Matrix Jz = contractDown(J, z); // Contracts J into p - 2 copies of z.
+  Vector Jzz = Jz * z;
+
+  double f = factorial(p);
+
+  Matrix hessian = ((p - 1) * p / f) * Jz;
+  Vector gradient = (p / f) * Jzz;
+  Scalar hamiltonian = (1 / f) * Jzz.dot(z);
+
+  return {hamiltonian, gradient, hessian};
+}
+
+std::tuple<double, Vector> WdW(const Tensor& J, const Vector& z) {
+  Vector gradient;
+  Matrix hessian;
+  std::tie(std::ignore, gradient, hessian) = hamGradHess(J, z);
+
+  Vector projectedGradient = gradient - (gradient.dot(z) / (double)z.size()) * z;
+
+  double W = projectedGradient.cwiseAbs2().sum();
+  Vector dW = hessian.conjugate() * projectedGradient;
+
+  return {W, dW};
+}
diff --git a/stereographic.hpp b/stereographic.hpp
index 7dd871c..4109bc7 100644
--- a/stereographic.hpp
+++ b/stereographic.hpp
@@ -1,3 +1,6 @@
+#include <eigen3/Eigen/Cholesky>
+
+#include "p-spin.hpp"
 
 Vector stereographicToEuclidean(const Vector& ζ) {
   unsigned N = ζ.size() + 1;
@@ -67,8 +70,7 @@ Matrix stereographicMetric(const Vector& ζ) {
   return g;
 }
 
-std::tuple<std::complex<double>, Vector, Matrix> stereographicHamGradHess(const Tensor3& J,
-                                                                          const Vector& ζ) {
+std::tuple<Scalar, Vector, Matrix> stereographicHamGradHess(const Tensor& J, const Vector& ζ) {
   Vector grad;
   Matrix hess;