big rearrangement of files to make libraries and example (research) files clearer, and changed to c++ std lib random numbers

47 files changed, 283 insertions, 575 deletions

diff --git a/.gitmodules b/.gitmodules
new file mode 100644
index 0000000..66f0577
--- /dev/null
+++ b/.gitmodules
@@ -0,0 +1,3 @@
+[submodule "examples/include/randutils"]
+	path = examples/include/randutils
+	url = https://gist.github.com/imneme/540829265469e673d045
diff --git a/CMakeLists.txt b/CMakeLists.txt
index 7bf604c..073ae14 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -1,6 +1,6 @@
 
-cmake_minimum_required(VERSION 3.0)
-project(wolff)
+cmake_minimum_required(VERSION 3.9)
+project(wolff DESCRIPTION "a library for preforming the wolff algorithm")
 
 set(CMAKE_CXX_FLAGS_DEBUG "-g -Wall")
 set(CMAKE_CXX_FLAGS_RELEASE "-O3")
@@ -8,89 +8,8 @@ set(CMAKE_CXX_FLAGS_RELEASE "-O3")
 set (CMAKE_CXX_STANDARD 17)
 set (CMAKE_C_STANDARD 11)
 
-include_directories(lib ~/.local/include)
-link_directories(~/.local/lib)
+include(GNUInstallDirs)
 
-file(GLOB CSOURCES lib/*.c)
-file(GLOB CPPSOURCES lib/*.cpp)
-
-add_executable(wolff_ising src/wolff_ising.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_ising_2D_0 src/wolff_ising.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_dgm src/wolff_dgm.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_cgm src/wolff_cgm.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_3potts src/wolff_potts.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_4potts src/wolff_potts.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_7potts src/wolff_potts.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_3clock src/wolff_clock.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_5clock src/wolff_clock.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_planar src/wolff_On.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_planar_2D src/wolff_On.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_planar_2D_0 src/wolff_On.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(wolff_heisenberg src/wolff_On.cpp ${CPPSOURCES} ${CSOURCES})
-add_executable(analyze_correlations src/analyze_correlations.cpp ${CPPSOURCES} ${CSOURCES})
-
-SET_TARGET_PROPERTIES(wolff_ising_2D_0 PROPERTIES COMPILE_FLAGS "-DDIMENSION=2 -DNOFIELD")
-SET_TARGET_PROPERTIES(wolff_3potts PROPERTIES COMPILE_FLAGS "-DPOTTSQ=3")
-SET_TARGET_PROPERTIES(wolff_4potts PROPERTIES COMPILE_FLAGS "-DPOTTSQ=4")
-SET_TARGET_PROPERTIES(wolff_7potts PROPERTIES COMPILE_FLAGS "-DPOTTSQ=7")
-SET_TARGET_PROPERTIES(wolff_3clock PROPERTIES COMPILE_FLAGS "-DPOTTSQ=3")
-SET_TARGET_PROPERTIES(wolff_5clock PROPERTIES COMPILE_FLAGS "-DPOTTSQ=5")
-SET_TARGET_PROPERTIES(wolff_planar PROPERTIES COMPILE_FLAGS "-DN_COMP=2")
-SET_TARGET_PROPERTIES(wolff_planar_2D PROPERTIES COMPILE_FLAGS "-DN_COMP=2 -DDIMENSION=2")
-SET_TARGET_PROPERTIES(wolff_planar_2D_0 PROPERTIES COMPILE_FLAGS "-DN_COMP=2 -DDIMENSION=2 -DNOFIELD")
-SET_TARGET_PROPERTIES(wolff_heisenberg PROPERTIES COMPILE_FLAGS "-DN_COMP=3")
-
-find_library(GSL REQUIRED NAMES gsl)
-find_library(FFTW REQUIRED NAMES fftw3)
-find_library(M REQUIRED NAMES m)
-FIND_LIBRARY(GL NAMES GL)
-FIND_LIBRARY(GLU NAMES GLU)
-FIND_LIBRARY(GLUT NAMES glut)
-
-target_link_libraries(analyze_correlations cblas gsl fftw3 m)
-
-if (${GLUT} MATCHES "GLUT-NOTFOUND")
-  target_link_libraries(wolff_ising cblas gsl m)
-  target_link_libraries(wolff_ising_2D_0 cblas gsl m)
-  target_link_libraries(wolff_dgm cblas gsl m)
-  target_link_libraries(wolff_cgm cblas gsl m)
-  target_link_libraries(wolff_3potts cblas gsl m)
-  target_link_libraries(wolff_4potts cblas gsl m)
-  target_link_libraries(wolff_7potts cblas gsl m)
-  target_link_libraries(wolff_3clock cblas gsl m)
-  target_link_libraries(wolff_5clock cblas gsl m)
-  target_link_libraries(wolff_heisenberg cblas gsl m)
-  target_link_libraries(wolff_planar cblas gsl m)
-  target_link_libraries(wolff_planar_2D cblas gsl m)
-  target_link_libraries(wolff_planar_2D_0 cblas gsl m)
-else()
-  target_link_libraries(wolff_ising cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_ising_2D_0 cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_dgm cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_cgm cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_3potts cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_4potts cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_7potts cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_3clock cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_5clock cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_heisenberg cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_planar cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_planar_2D cblas gsl m glut GL GLU)
-  target_link_libraries(wolff_planar_2D_0 cblas gsl m glut GL GLU)
-  target_compile_definitions(wolff_ising PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_ising_2D_0 PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_dgm PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_cgm PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_3potts PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_4potts PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_7potts PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_3clock PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_5clock PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_planar PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_planar_2D PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_planar_2D_0 PUBLIC HAVE_GLUT)
-  target_compile_definitions(wolff_heisenberg PUBLIC HAVE_GLUT)
-endif()
-
-install(TARGETS wolff_ising wolff_dgm wolff_cgm wolff_3potts wolff_4potts wolff_3clock wolff_heisenberg wolff_planar analyze_correlations DESTINATION bin)
+add_subdirectory(lib)
+add_subdirectory(examples)
 
diff --git a/examples/CMakeLists.txt b/examples/CMakeLists.txt
new file mode 100644
index 0000000..baf0a10
--- /dev/null
+++ b/examples/CMakeLists.txt
@@ -0,0 +1,7 @@
+
+add_library(wolff_examples INTERFACE)
+
+target_include_directories(wolff_examples INTERFACE include)
+
+add_subdirectory(src)
+
diff --git a/lib/colors.h b/examples/include/colors.h
index 04d39a8..abf137c 100644
--- a/lib/colors.h
+++ b/examples/include/colors.h
@@ -1,6 +1,6 @@
 #pragma once
 
-#include "types.h"
+#include <wolff/types.h>
 
 double hue_to_R(double theta) {
   if (((M_PI / 3 <= theta) && (theta < 2 * M_PI / 3)) || ((4 * M_PI / 3 <= theta) && (theta < 5 * M_PI / 3))) {
diff --git a/lib/correlation.h b/examples/include/correlation.hpp
index 29357a5..042cff3 100644
--- a/lib/correlation.h
+++ b/examples/include/correlation.hpp
@@ -1,8 +1,8 @@
 
 #pragma once
 
-#include "types.h"
-#include "state.h"
+#include <wolff/types.h>
+#include <wolff/state.hpp>
 
 #include <fftw3.h>
 
diff --git a/lib/measure.h b/examples/include/measure.hpp
index 2c5ffb7..e20353c 100644
--- a/lib/measure.h
+++ b/examples/include/measure.hpp
@@ -1,8 +1,8 @@
 
 #pragma once
 
-#include "state.h"
-#include "correlation.h"
+#include <wolff/state.hpp>
+#include "correlation.hpp"
 #include <functional>
 
 #define POSSIBLE_MEASUREMENTS 4
diff --git a/examples/include/randutils b/examples/include/randutils
new file mode 160000
+Subproject 8486a610a954a8248c12485fb4cfc390a5f5f85
diff --git a/examples/src/CMakeLists.txt b/examples/src/CMakeLists.txt
new file mode 100644
index 0000000..3397426
--- /dev/null
+++ b/examples/src/CMakeLists.txt
@@ -0,0 +1,4 @@
+
+add_subdirectory(models)
+add_subdirectory(tools)
+
diff --git a/examples/src/models/CMakeLists.txt b/examples/src/models/CMakeLists.txt
new file mode 100644
index 0000000..0b0c111
--- /dev/null
+++ b/examples/src/models/CMakeLists.txt
@@ -0,0 +1,6 @@
+
+add_subdirectory(ising)
+add_subdirectory(On)
+add_subdirectory(potts)
+add_subdirectory(roughening)
+
diff --git a/examples/src/models/On/CMakeLists.txt b/examples/src/models/On/CMakeLists.txt
new file mode 100644
index 0000000..26985b9
--- /dev/null
+++ b/examples/src/models/On/CMakeLists.txt
@@ -0,0 +1,29 @@
+
+add_executable(wolff_planar               wolff_On.cpp)
+add_executable(wolff_planar_2d_no-field   wolff_On.cpp)
+add_executable(wolff_heisenberg           wolff_On.cpp)
+
+set_target_properties(wolff_planar             PROPERTIES COMPILE_FLAGS "-DN_COMP=2")
+set_target_properties(wolff_planar_2d_no-field PROPERTIES COMPILE_FLAGS "-DN_COMP=2 -DDIMENSION=2 -DNOFIELD")
+set_target_properties(wolff_heisenberg         PROPERTIES COMPILE_FLAGS "-DN_COMP=3")
+
+find_library(GL NAMES GL)
+find_library(GLU NAMES GLU)
+find_library(GLUT NAMES glut)
+
+if (${GLUT} MATCHES "GLUT-NOTFOUND")
+  target_link_libraries(wolff_planar             wolff wolff_examples)
+  target_link_libraries(wolff_planar_2d_no-field wolff wolff_examples)
+  target_link_libraries(wolff_heisenberg         wolff wolff_examples)
+else()
+  target_compile_definitions(wolff_planar             PUBLIC HAVE_GLUT)
+  target_compile_definitions(wolff_planar_2d_no-field             PUBLIC HAVE_GLUT)
+  target_compile_definitions(wolff_heisenberg             PUBLIC HAVE_GLUT)
+
+  target_link_libraries(wolff_planar             wolff wolff_examples glut GL GLU)
+  target_link_libraries(wolff_planar_2d_no-field wolff wolff_examples glut GL GLU)
+  target_link_libraries(wolff_heisenberg             wolff wolff_examples glut GL GLU)
+endif()
+
+install(TARGETS wolff_planar wolff_planar_2d_no-field wolff_heisenberg DESTINATION ${CMAKE_INSTALL_BINDIR})
+
diff --git a/lib/orthogonal.h b/examples/src/models/On/orthogonal.hpp
index e1bf33a..f13357f 100644
--- a/lib/orthogonal.h
+++ b/examples/src/models/On/orthogonal.hpp
@@ -2,13 +2,12 @@
 #pragma once
 
 #include <stdlib.h>
-#include <gsl/gsl_rng.h>
-#include <gsl/gsl_randist.h>
+#include <random>
 #include <cmath>
 
-#include "state.h"
-#include "types.h"
-#include "vector.h"
+#include <wolff/state.hpp>
+#include <wolff/types.h>
+#include "vector.hpp"
 
 template <q_t q, class T>
 class orthogonal_t : public std::array<std::array<T, q>, q> {
@@ -117,14 +116,15 @@ class orthogonal_t : public std::array<std::array<T, q>, q> {
 
 
 template <q_t q>
-orthogonal_t <q, double> generate_rotation_uniform (gsl_rng *r, const vector_t <q, double>& v) {
+orthogonal_t <q, double> generate_rotation_uniform (std::mt19937& r, const vector_t <q, double>& v) {
+  std::normal_distribution<double> dist(0.0,1.0);
   orthogonal_t <q, double> ptr;
   ptr.is_reflection = true;
 
   double v2 = 0;
 
   for (q_t i = 0; i < q; i++) {
-    ptr[0][i] = gsl_ran_ugaussian(r);
+    ptr[0][i] = dist(r);
     v2 += ptr[0][i] * ptr[0][i];
   }
 
@@ -138,14 +138,16 @@ orthogonal_t <q, double> generate_rotation_uniform (gsl_rng *r, const vector_t <
 }
 
 template <q_t q>
-orthogonal_t <q, double> generate_rotation_perturbation (gsl_rng *r, const vector_t <q, double>& v0, double epsilon, unsigned int n) {
+orthogonal_t <q, double> generate_rotation_perturbation (std::mt19937& r, const vector_t <q, double>& v0, double epsilon, unsigned int n) {
+  std::normal_distribution<double> dist(0.0,1.0);
   orthogonal_t <q, double> m;
   m.is_reflection = true;
 
   vector_t <q, double> v;
 
   if (n > 1) {
-    unsigned int rotation = gsl_rng_uniform_int(r, n);
+    std::uniform_int_distribution<unsigned int> udist(0, n);
+    unsigned int rotation = udist(r);
 
     double cosr = cos(2 * M_PI * rotation / (double)n / 2.0);
     double sinr = sin(2 * M_PI * rotation / (double)n / 2.0);
@@ -163,7 +165,7 @@ orthogonal_t <q, double> generate_rotation_perturbation (gsl_rng *r, const vecto
   double m_dot_v = 0;
 
   for (q_t i = 0; i < q; i++) {
-    m[0][i] = gsl_ran_ugaussian(r); // create a random vector
+    m[0][i] = dist(r); // create a random vector
     m_dot_v += m[0][i] * v[i];
   }
 
@@ -182,7 +184,7 @@ orthogonal_t <q, double> generate_rotation_perturbation (gsl_rng *r, const vecto
 
   v2 = 0;
 
-  double factor = gsl_ran_gaussian(r, epsilon);
+  double factor = epsilon * dist(r);
 
   for (q_t i = 0; i < q; i++) {
     m[0][i] += factor * v[i]; // perturb orthogonal vector in original direction
diff --git a/lib/vector.h b/examples/src/models/On/vector.hpp
index 7d0ee36..1cdb60a 100644
--- a/lib/vector.h
+++ b/examples/src/models/On/vector.hpp
@@ -5,7 +5,7 @@
 #include <cmath>
 #include <array>
 
-#include "types.h"
+#include <wolff/types.h>
 
 template <q_t q, class T>
 class vector_t : public std::array<T, q> {
diff --git a/src/wolff_On.cpp b/examples/src/models/On/wolff_On.cpp
index f6661af..e3568c7 100644
--- a/src/wolff_On.cpp
+++ b/examples/src/models/On/wolff_On.cpp
@@ -6,13 +6,13 @@
 #include <GL/glut.h>
 #endif
 
-#include <orthogonal.h>
-#include <vector.h>
+#include "orthogonal.hpp"
+#include "vector.hpp"
 
-#include <wolff.h>
-#include <measure.h>
+#include <wolff.hpp>
+#include <measure.hpp>
 #include <colors.h>
-#include <rand.h>
+#include <randutils/randutils.hpp>
 
 typedef orthogonal_t <N_COMP, double> orthogonal_R_t;
 typedef vector_t <N_COMP, double> vector_R_t;
@@ -139,7 +139,7 @@ int main(int argc, char *argv[]) {
 
   const char *pert_type;
 
-  std::function <orthogonal_R_t(gsl_rng *, vector_R_t)> gen_R;
+  std::function <orthogonal_R_t(std::mt19937&, vector_R_t)> gen_R;
 
   if (use_pert) {
     double Hish;
@@ -238,8 +238,8 @@ int main(int argc, char *argv[]) {
   }
 
   // initialize random number generator
-  gsl_rng *r = gsl_rng_alloc(gsl_rng_taus2);
-  gsl_rng_set(r, rand_seed());
+  randutils::auto_seed_128 seeds;
+  std::mt19937 rng{seeds};
 
 #ifndef NOFIELD
   state_t <orthogonal_R_t, vector_R_t> s(D, L, T, dot <N_COMP, double>, H);
@@ -252,17 +252,16 @@ int main(int argc, char *argv[]) {
     printf("\n");
     while (sum_of_clusterSize < N * s.nv) {
       printf("\033[F\033[J\033[F\033[JWOLFF: sweep %" PRIu64 " / %" PRIu64 ": E = %.2f, S = %" PRIv "\n", (count_t)((double)sum_of_clusterSize / (double)s.nv), N, s.E, s.last_cluster_size);
-      wolff <orthogonal_R_t, vector_R_t> (N, s, gen_R, measurements, r, silent);
+      wolff <orthogonal_R_t, vector_R_t> (N, s, gen_R, measurements, rng, silent);
       N_rounds++;
     }
     printf("\033[F\033[J\033[F\033[JWOLFF: sweep %" PRIu64 " / %" PRIu64 ": E = %.2f, S = %" PRIv "\n\n", (count_t)((double)sum_of_clusterSize / (double)s.nv), N, s.E, s.last_cluster_size);
   } else {
-    wolff <orthogonal_R_t, vector_R_t> (N, s, gen_R, measurements, r, silent);
+    wolff <orthogonal_R_t, vector_R_t> (N, s, gen_R, measurements, rng, silent);
   }
 
   measure_free_files(measurement_flags, outfiles);
   free(H_vec);
-  gsl_rng_free(r);
 
   return 0;
 }
diff --git a/examples/src/models/ising/CMakeLists.txt b/examples/src/models/ising/CMakeLists.txt
new file mode 100644
index 0000000..e8fbc9a
--- /dev/null
+++ b/examples/src/models/ising/CMakeLists.txt
@@ -0,0 +1,28 @@
+
+add_executable(wolff_ising             wolff_ising.cpp)
+add_executable(wolff_ising_2d          wolff_ising.cpp)
+add_executable(wolff_ising_2d_no-field wolff_ising.cpp)
+
+set_target_properties(wolff_ising_2d          PROPERTIES COMPILE_FLAGS "-DDIMENSION=2")
+set_target_properties(wolff_ising_2d_no-field PROPERTIES COMPILE_FLAGS "-DDIMENSION=2 -DNOFIELD")
+
+find_library(GL NAMES GL)
+find_library(GLU NAMES GLU)
+find_library(GLUT NAMES glut)
+
+if (${GLUT} MATCHES "GLUT-NOTFOUND")
+  target_link_libraries(wolff_ising             wolff wolff_examples)
+  target_link_libraries(wolff_ising_2d          wolff wolff_examples)
+  target_link_libraries(wolff_ising_2d_no-field wolff wolff_examples)
+else()
+  target_compile_definitions(wolff_ising             PUBLIC HAVE_GLUT)
+  target_compile_definitions(wolff_ising_2d          PUBLIC HAVE_GLUT)
+  target_compile_definitions(wolff_ising_2d_no-field PUBLIC HAVE_GLUT)
+
+  target_link_libraries(wolff_ising             wolff wolff_examples glut GL GLU)
+  target_link_libraries(wolff_ising_2d          wolff wolff_examples glut GL GLU)
+  target_link_libraries(wolff_ising_2d_no-field wolff wolff_examples glut GL GLU)
+endif()
+
+install(TARGETS wolff_ising wolff_ising_2d wolff_ising_2d_no-field DESTINATION ${CMAKE_INSTALL_BINDIR})
+
diff --git a/lib/ising.h b/examples/src/models/ising/ising.hpp
index 45058fb..ae20840 100644
--- a/lib/ising.h
+++ b/examples/src/models/ising/ising.hpp
@@ -3,7 +3,7 @@
 #include <cmath>
 #include <stdio.h>
 
-#include "types.h"
+#include <wolff/types.h>
 
 class ising_t {
   public:
diff --git a/src/wolff_ising.cpp b/examples/src/models/ising/wolff_ising.cpp
index a6f43b1..5bdaa82 100644
--- a/src/wolff_ising.cpp
+++ b/examples/src/models/ising/wolff_ising.cpp
@@ -8,21 +8,20 @@
 #endif
 
 // include your group and spin space
-#include <z2.h>
-#include <ising.h>
+#include "z2.hpp"
+#include "ising.hpp"
 
 // finite_states.h can be included for spin types that have special variables
 // defined, and it causes wolff execution to use precomputed bond probabilities
-#include <finite_states.h>
+#include <wolff/finite_states.hpp>
 
-// rand.h uses a unix-specific way to seed the random number generator
-#include <rand.h>
+#include <randutils/randutils.hpp>
 
-// measure.h contains useful functions for saving timeseries to files
-#include <measure.h>
+// measure.hpp contains useful functions for saving timeseries to files
+#include <measure.hpp>
 
-// include wolff.h
-#include <wolff.h>
+// include wolff.hpp
+#include <wolff.hpp>
 
 int main(int argc, char *argv[]) {
 
@@ -95,8 +94,8 @@ int main(int argc, char *argv[]) {
   }
 
   // initialize random number generator
-  gsl_rng *r = gsl_rng_alloc(gsl_rng_taus2);
-  gsl_rng_set(r, rand_seed());
+  randutils::auto_seed_128 seeds;
+  std::mt19937 rng{seeds};
 
   // define spin-spin coupling
   std::function <double(const ising_t&, const ising_t&)> Z = [] (const ising_t& s1, const ising_t& s2) -> double {
@@ -124,7 +123,7 @@ int main(int argc, char *argv[]) {
 #endif
 
   // define function that generates self-inverse rotations
-  std::function <z2_t(gsl_rng *, ising_t)> gen_R = [] (gsl_rng *, const ising_t& s) -> z2_t {
+  std::function <z2_t(std::mt19937&, ising_t)> gen_R = [] (std::mt19937&, const ising_t& s) -> z2_t {
     return z2_t(true);
   };
 
@@ -183,16 +182,14 @@ int main(int argc, char *argv[]) {
     printf("\n");
     while (sum_of_clusterSize < N * s.nv) {
       printf("\033[F\033[J\033[F\033[JWOLFF: sweep %" PRIu64 " / %" PRIu64 ": E = %.2f, S = %" PRIv "\n", (count_t)((double)sum_of_clusterSize / (double)s.nv), N, s.E, s.last_cluster_size);
-      wolff(N, s, gen_R, measurements, r, silent);
+      wolff(N, s, gen_R, measurements, rng, silent);
       N_rounds++;
     }
     printf("\033[F\033[J\033[F\033[JWOLFF: sweep %" PRIu64 " / %" PRIu64 ": E = %.2f, S = %" PRIv "\n\n", (count_t)((double)sum_of_clusterSize / (double)s.nv), N, s.E, s.last_cluster_size);
   } else {
-    wolff(N, s, gen_R, measurements, r, silent);
+    wolff(N, s, gen_R, measurements, rng, silent);
   }
 
-  // free the random number generator
-  gsl_rng_free(r);
   measure_free_files(measurement_flags, outfiles);
 
   return 0;
diff --git a/lib/z2.h b/examples/src/models/ising/z2.hpp
index a18d740..19b6c05 100644
--- a/lib/z2.h
+++ b/examples/src/models/ising/z2.hpp
@@ -1,8 +1,8 @@
 
 #pragma once
 
-#include "types.h"
-#include "ising.h"
+#include <wolff/types.h>
+#include "ising.hpp"
 
 /* The minimum definition for a group type R_t to act on a spin type X_t is
  * given by the following.
diff --git a/examples/src/models/potts/CMakeLists.txt b/examples/src/models/potts/CMakeLists.txt
new file mode 100644
index 0000000..53133b9
--- /dev/null
+++ b/examples/src/models/potts/CMakeLists.txt
@@ -0,0 +1,39 @@
+
+add_executable(wolff_3potts             wolff_potts.cpp)
+add_executable(wolff_4potts             wolff_potts.cpp)
+add_executable(wolff_7potts             wolff_potts.cpp)
+add_executable(wolff_3clock             wolff_clock.cpp)
+add_executable(wolff_5clock             wolff_clock.cpp)
+
+set_target_properties(wolff_3potts          PROPERTIES COMPILE_FLAGS "-DPOTTSQ=3")
+set_target_properties(wolff_4potts          PROPERTIES COMPILE_FLAGS "-DPOTTSQ=4")
+set_target_properties(wolff_7potts          PROPERTIES COMPILE_FLAGS "-DPOTTSQ=7")
+set_target_properties(wolff_3clock          PROPERTIES COMPILE_FLAGS "-DPOTTSQ=3")
+set_target_properties(wolff_5clock          PROPERTIES COMPILE_FLAGS "-DPOTTSQ=5")
+
+find_library(GL NAMES GL)
+find_library(GLU NAMES GLU)
+find_library(GLUT NAMES glut)
+
+if (${GLUT} MATCHES "GLUT-NOTFOUND")
+  target_link_libraries(wolff_3potts wolff wolff_examples)
+  target_link_libraries(wolff_4potts wolff wolff_examples)
+  target_link_libraries(wolff_7potts wolff wolff_examples)
+  target_link_libraries(wolff_3clock wolff wolff_examples)
+  target_link_libraries(wolff_5clock wolff wolff_examples)
+else()
+  target_compile_definitions(wolff_3potts PUBLIC HAVE_GLUT)
+  target_compile_definitions(wolff_4potts PUBLIC HAVE_GLUT)
+  target_compile_definitions(wolff_7potts PUBLIC HAVE_GLUT)
+  target_compile_definitions(wolff_3clock PUBLIC HAVE_GLUT)
+  target_compile_definitions(wolff_5clock PUBLIC HAVE_GLUT)
+
+  target_link_libraries(wolff_3potts             wolff wolff_examples glut GL GLU)
+  target_link_libraries(wolff_4potts             wolff wolff_examples glut GL GLU)
+  target_link_libraries(wolff_7potts             wolff wolff_examples glut GL GLU)
+  target_link_libraries(wolff_3clock             wolff wolff_examples glut GL GLU)
+  target_link_libraries(wolff_5clock             wolff wolff_examples glut GL GLU)
+endif()
+
+install(TARGETS wolff_3potts wolff_4potts wolff_7potts wolff_3clock wolff_5clock DESTINATION ${CMAKE_INSTALL_BINDIR})
+
diff --git a/lib/dihedral.h b/examples/src/models/potts/dihedral.hpp
index 8d0472b..cbc5687 100644
--- a/lib/dihedral.h
+++ b/examples/src/models/potts/dihedral.hpp
@@ -1,8 +1,8 @@
 
 #pragma once
 
-#include "types.h"
-#include "potts.h"
+#include <wolff/types.h>
+#include "potts.hpp"
 
 template <q_t q>
 class dihedral_t {
diff --git a/lib/potts.h b/examples/src/models/potts/potts.hpp
index 1764e53..f4765e2 100644
--- a/lib/potts.h
+++ b/examples/src/models/potts/potts.hpp
@@ -3,8 +3,8 @@
 #include <cmath>
 #include <stdio.h>
 
-#include "types.h"
-#include "vector.h"
+#include <wolff/types.h>
+#include "../On/vector.hpp"
 
 template <q_t q>
 class potts_t {
diff --git a/lib/symmetric.h b/examples/src/models/potts/symmetric.hpp
index 9e9b6e4..8636f15 100644
--- a/lib/symmetric.h
+++ b/examples/src/models/potts/symmetric.hpp
@@ -3,8 +3,8 @@
 
 #include <stdlib.h>
 #include <array>
-#include "types.h"
-#include "potts.h"
+#include <wolff/types.h>
+#include "potts.hpp"
 
 template <q_t q>
 class symmetric_t : public std::array<q_t, q> {
diff --git a/src/wolff_clock.cpp b/examples/src/models/potts/wolff_clock.cpp
index 3dec284..020415d 100644
--- a/src/wolff_clock.cpp
+++ b/examples/src/models/potts/wolff_clock.cpp
@@ -6,17 +6,18 @@
 #endif
 
 // include your group and spin space
-#include <dihedral.h>
-#include <potts.h>
+#include "dihedral.hpp"
+#include "potts.hpp"
 #include <colors.h>
 
 // hack to speed things up considerably
 #define N_STATES POTTSQ
-#include <finite_states.h>
+#include <wolff/finite_states.hpp>
 
-// include wolff.h
-#include <rand.h>
-#include <wolff.h>
+#include <randutils/randutils.hpp>
+
+// include wolff.hpp
+#include <wolff.hpp>
 
 typedef state_t <dihedral_t<POTTSQ>, potts_t<POTTSQ>> sim_t;
 
@@ -74,8 +75,8 @@ int main(int argc, char *argv[]) {
   }
 
   // initialize random number generator
-  gsl_rng *r = gsl_rng_alloc(gsl_rng_mt19937);
-  gsl_rng_set(r, rand_seed());
+  randutils::auto_seed_128 seeds;
+  std::mt19937 rng{seeds};
 
   // define spin-spin coupling
   std::function <double(const potts_t<POTTSQ>&, const potts_t<POTTSQ>&)> Z = [] (const potts_t<POTTSQ>& s1, const potts_t<POTTSQ>& s2) -> double {
@@ -91,10 +92,11 @@ int main(int argc, char *argv[]) {
   state_t <dihedral_t<POTTSQ>, potts_t<POTTSQ>> s(D, L, T, Z, B);
 
   // define function that generates self-inverse rotations
-  std::function <dihedral_t<POTTSQ>(gsl_rng *, potts_t<POTTSQ>)> gen_R = [] (gsl_rng *r, potts_t<POTTSQ> v) -> dihedral_t<POTTSQ> {
+  std::function <dihedral_t<POTTSQ>(std::mt19937&, potts_t<POTTSQ>)> gen_R = [] (std::mt19937& r, potts_t<POTTSQ> v) -> dihedral_t<POTTSQ> {
     dihedral_t<POTTSQ> rot;
     rot.is_reflection = true;
-    q_t x = gsl_rng_uniform_int(r, POTTSQ - 1);
+    std::uniform_int_distribution<q_t> dist(0, POTTSQ - 1);
+    q_t x = dist(r);
     rot.x = (2 * v.x + x + 1) % POTTSQ;
 
     return rot;
@@ -137,13 +139,12 @@ int main(int argc, char *argv[]) {
   }
 
   // run wolff for N cluster flips
-  wolff(N, s, gen_R, measurement, r, silent);
+  wolff(N, s, gen_R, measurement, rng, silent);
 
   // tell us what we found!
   printf("%" PRIcount " %d-Potts runs completed. D = %" PRID ", L = %" PRIL ", T = %g, H = %g, <M> = %g\n", N, POTTSQ, D, L, T, H_vec[0], average_M);
 
   // free the random number generator
-  gsl_rng_free(r);
 
   if (draw) {
   }
diff --git a/src/wolff_potts.cpp b/examples/src/models/potts/wolff_potts.cpp
index 9fe3ffe..a1e9284 100644
--- a/src/wolff_potts.cpp
+++ b/examples/src/models/potts/wolff_potts.cpp
@@ -7,18 +7,18 @@
 #endif
 
 // include your group and spin space
-#include <symmetric.h>
-#include <potts.h>
+#include "symmetric.hpp"
+#include "potts.hpp"
 
 // hack to speed things up considerably
 #define N_STATES POTTSQ
-#include <finite_states.h>
+#include <wolff/finite_states.hpp>
 
 // include wolff.h
-#include <measure.h>
+#include <measure.hpp>
 #include <colors.h>
-#include <rand.h>
-#include <wolff.h>
+#include <randutils/randutils.hpp>
+#include <wolff.hpp>
 
 typedef state_t <symmetric_t<POTTSQ>, potts_t<POTTSQ>> sim_t;
 
@@ -95,8 +95,8 @@ int main(int argc, char *argv[]) {
   }
 
   // initialize random number generator
-  gsl_rng *r = gsl_rng_alloc(gsl_rng_mt19937);
-  gsl_rng_set(r, rand_seed());
+  randutils::auto_seed_128 seeds;
+  std::mt19937 rng{seeds};
 
   // define spin-spin coupling
   std::function <double(const potts_t<POTTSQ>&, const potts_t<POTTSQ>&)> Z = [] (const potts_t<POTTSQ>& s1, const potts_t<POTTSQ>& s2) -> double {
@@ -116,10 +116,11 @@ int main(int argc, char *argv[]) {
   state_t <symmetric_t<POTTSQ>, potts_t<POTTSQ>> s(D, L, T, Z, B);
 
   // define function that generates self-inverse rotations
-  std::function <symmetric_t<POTTSQ>(gsl_rng *, potts_t<POTTSQ>)> gen_R = [] (gsl_rng *r, potts_t<POTTSQ> v) -> symmetric_t<POTTSQ> {
+  std::function <symmetric_t<POTTSQ>(std::mt19937&, potts_t<POTTSQ>)> gen_R = [] (std::mt19937& r, potts_t<POTTSQ> v) -> symmetric_t<POTTSQ> {
     symmetric_t<POTTSQ> rot;
 
-    q_t j = gsl_rng_uniform_int(r, POTTSQ - 1);
+    std::uniform_int_distribution<q_t> dist(0, POTTSQ - 1);
+    q_t j = dist(r);
     q_t swap_v;
     if (j < v.x) {
       swap_v = j;
@@ -194,16 +195,15 @@ int main(int argc, char *argv[]) {
     printf("\n");
     while (sum_of_clusterSize < N * s.nv) {
       printf("\033[F\033[J\033[F\033[JWOLFF: sweep %" PRIu64 " / %" PRIu64 ": E = %.2f, S = %" PRIv "\n", (count_t)((double)sum_of_clusterSize / (double)s.nv), N, s.E, s.last_cluster_size);
-      wolff(N, s, gen_R, measurements, r, silent);
+      wolff(N, s, gen_R, measurements, rng, silent);
       N_rounds++;
     }
     printf("\033[F\033[J\033[F\033[JWOLFF: sweep %" PRIu64 " / %" PRIu64 ": E = %.2f, S = %" PRIv "\n\n", (count_t)((double)sum_of_clusterSize / (double)s.nv), N, s.E, s.last_cluster_size);
   } else {
-    wolff(N, s, gen_R, measurements, r, silent);
+    wolff(N, s, gen_R, measurements, rng, silent);
   }
 
   // free the random number generator
-  gsl_rng_free(r);
   free(H_vec);
   measure_free_files(measurement_flags, outfiles);
 
diff --git a/examples/src/models/roughening/CMakeLists.txt b/examples/src/models/roughening/CMakeLists.txt
new file mode 100644
index 0000000..51a8644
--- /dev/null
+++ b/examples/src/models/roughening/CMakeLists.txt
@@ -0,0 +1,21 @@
+
+add_executable(wolff_dgm             wolff_dgm.cpp)
+add_executable(wolff_cgm             wolff_cgm.cpp)
+
+find_library(GL NAMES GL)
+find_library(GLU NAMES GLU)
+find_library(GLUT NAMES glut)
+
+if (${GLUT} MATCHES "GLUT-NOTFOUND")
+  target_link_libraries(wolff_dgm             wolff wolff_examples)
+  target_link_libraries(wolff_cgm             wolff wolff_examples)
+else()
+  target_compile_definitions(wolff_dgm             PUBLIC HAVE_GLUT)
+  target_compile_definitions(wolff_cgm             PUBLIC HAVE_GLUT)
+
+  target_link_libraries(wolff_dgm             wolff wolff_examples glut GL GLU)
+  target_link_libraries(wolff_cgm             wolff wolff_examples glut GL GLU)
+endif()
+
+install(TARGETS wolff_dgm wolff_cgm DESTINATION ${CMAKE_INSTALL_BINDIR})
+
diff --git a/lib/dihedral_inf.h b/examples/src/models/roughening/dihedral_inf.hpp
index a064b48..19fa195 100644
--- a/lib/dihedral_inf.h
+++ b/examples/src/models/roughening/dihedral_inf.hpp
@@ -1,7 +1,7 @@
 
-#include "types.h"
+#include <wolff/types.h>
 #include <cmath>
-#include "height.h"
+#include "height.hpp"
 
 template <class T>
 class dihedral_inf_t {
diff --git a/lib/height.h b/examples/src/models/roughening/height.hpp
index d2173fe..4023063 100644
--- a/lib/height.h
+++ b/examples/src/models/roughening/height.hpp
@@ -4,7 +4,7 @@
 #include <cmath>
 #include <stdio.h>
 
-#include "types.h"
+#include <wolff/types.h>
 
 /* The following is the minimum definition of a spin class.
  *
diff --git a/src/wolff_cgm.cpp b/examples/src/models/roughening/wolff_cgm.cpp
index ce91bf2..65f8d66 100644
--- a/src/wolff_cgm.cpp
+++ b/examples/src/models/roughening/wolff_cgm.cpp
@@ -6,12 +6,13 @@
 #endif
 
 // include your group and spin space
-#include <dihedral_inf.h>
-#include <height.h>
+#include "dihedral_inf.hpp"
+#include "height.hpp"
+
+#include <randutils/randutils.hpp>
 
 // include wolff.h
-#include <rand.h>
-#include <wolff.h>
+#include <wolff.hpp>
 
 typedef state_t <dihedral_inf_t<double>, height_t<double>> sim_t;
 
@@ -71,8 +72,8 @@ int main(int argc, char *argv[]) {
   }
 
   // initialize random number generator
-  gsl_rng *r = gsl_rng_alloc(gsl_rng_mt19937);
-  gsl_rng_set(r, rand_seed());
+  randutils::auto_seed_128 seeds;
+  std::mt19937 rng{seeds};
 
   // define spin-spin coupling
   std::function <double(const height_t<double>&, const height_t<double>&)> Z = [] (const height_t<double>& h1, const height_t<double>& h2) -> double {
@@ -88,11 +89,12 @@ int main(int argc, char *argv[]) {
   sim_t s(D, L, T, Z, B);
 
   // define function that generates self-inverse rotations
-  std::function <dihedral_inf_t<double>(gsl_rng *, height_t<double>)> gen_R = [=] (gsl_rng *r, height_t<double> h) -> dihedral_inf_t<double> {
+  std::function <dihedral_inf_t<double>(std::mt19937&, height_t<double>)> gen_R = [=] (std::mt19937& r, height_t<double> h) -> dihedral_inf_t<double> {
     dihedral_inf_t<double> rot;
     rot.is_reflection = true;
+    std::normal_distribution<double> dist(0.0,1.0);
 
-    double amount = epsilon * gsl_ran_ugaussian(r);
+    double amount = epsilon * dist(r);
 
     rot.x = 2 * h.x + amount;
 
@@ -149,13 +151,12 @@ int main(int argc, char *argv[]) {
   }
 
   // run wolff for N cluster flips
-  wolff(N, s, gen_R, measurement, r, silent);
+  wolff(N, s, gen_R, measurement, rng, silent);
 
   // tell us what we found!
   printf("%" PRIcount " DGM runs completed. D = %" PRID ", L = %" PRIL ", T = %g, H = %g, <M> = %g\n", N, D, L, T, H, average_M);
 
   // free the random number generator
-  gsl_rng_free(r);
 
   if (draw) {
   }
diff --git a/src/wolff_dgm.cpp b/examples/src/models/roughening/wolff_dgm.cpp
index 8667fb5..8395382 100644
--- a/src/wolff_dgm.cpp
+++ b/examples/src/models/roughening/wolff_dgm.cpp
@@ -6,12 +6,13 @@
 #endif
 
 // include your group and spin space
-#include <dihedral_inf.h>
-#include <height.h>
+#include "dihedral_inf.hpp"
+#include "height.hpp"
+
+#include <randutils/randutils.hpp>
 
 // include wolff.h
-#include <rand.h>
-#include <wolff.h>
+#include <wolff.hpp>
 
 typedef state_t <dihedral_inf_t<int64_t>, height_t<int64_t>> sim_t;
 
@@ -71,8 +72,8 @@ int main(int argc, char *argv[]) {
   }
 
   // initialize random number generator
-  gsl_rng *r = gsl_rng_alloc(gsl_rng_mt19937);
-  gsl_rng_set(r, rand_seed());
+  randutils::auto_seed_128 seeds;
+  std::mt19937 rng{seeds};
 
   // define spin-spin coupling
   std::function <double(const height_t<int64_t>&, const height_t<int64_t>&)> Z = [] (const height_t<int64_t>& h1, const height_t<int64_t>& h2) -> double {
@@ -88,18 +89,13 @@ int main(int argc, char *argv[]) {
   sim_t s(D, L, T, Z, B);
 
   // define function that generates self-inverse rotations
-  std::function <dihedral_inf_t<int64_t>(gsl_rng *, height_t<int64_t>)> gen_R = [=] (gsl_rng *r, height_t<int64_t> h) -> dihedral_inf_t<int64_t> {
+  std::function <dihedral_inf_t<int64_t>(std::mt19937&, height_t<int64_t>)> gen_R = [=] (std::mt19937& r, height_t<int64_t> h) -> dihedral_inf_t<int64_t> {
     dihedral_inf_t<int64_t> rot;
     rot.is_reflection = true;
 
-    int direction = gsl_rng_uniform_int(r, 2);
-    int64_t amount = gsl_rng_uniform_int(r, epsilon);
+    std::uniform_int_distribution<int64_t> dist(-epsilon,epsilon);
 
-    if (direction == 0) {
-      rot.x = 2 * h.x + (1 + amount);
-    } else {
-      rot.x = 2 * h.x - (1 + amount);
-    }
+    rot.x = 2 * h.x + dist(r);
 
     return rot;
   };
@@ -154,14 +150,11 @@ int main(int argc, char *argv[]) {
   }
 
   // run wolff for N cluster flips
-  wolff(N, s, gen_R, measurement, r, silent);
+  wolff(N, s, gen_R, measurement, rng, silent);
 
   // tell us what we found!
   printf("%" PRIcount " DGM runs completed. D = %" PRID ", L = %" PRIL ", T = %g, H = %g, <M> = %g\n", N, D, L, T, H, average_M);
 
-  // free the random number generator
-  gsl_rng_free(r);
-
   if (draw) {
   }
 
diff --git a/examples/src/tools/CMakeLists.txt b/examples/src/tools/CMakeLists.txt
new file mode 100644
index 0000000..1c73c2d
--- /dev/null
+++ b/examples/src/tools/CMakeLists.txt
@@ -0,0 +1,7 @@
+
+find_library(fftw REQUIRED NAMES fftw3)
+
+add_executable(analyze_correlations analyze_correlations.cpp)
+
+target_link_libraries(analyze_correlations fftw3 wolff)
+
diff --git a/src/analyze_correlations.cpp b/examples/src/tools/analyze_correlations.cpp
index 48ee426..abeaff3 100644
--- a/src/analyze_correlations.cpp
+++ b/examples/src/tools/analyze_correlations.cpp
@@ -1,5 +1,5 @@
 
-#include <types.h>
+#include <wolff/types.h>
 #include <cmath>
 #include <cstring>
 #include <stdio.h>
diff --git a/lib/CMakeLists.txt b/lib/CMakeLists.txt
new file mode 100644
index 0000000..a31500e
--- /dev/null
+++ b/lib/CMakeLists.txt
@@ -0,0 +1,17 @@
+
+project(libwolff LANGUAGES C CXX)
+
+add_library(wolff SHARED src/graph.cpp)
+
+target_include_directories(wolff PUBLIC $<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/include>
+    $<INSTALL_INTERFACE:include>)
+
+install(TARGETS wolff EXPORT wolffConfig
+    ARCHIVE  DESTINATION ${CMAKE_INSTALL_LIBDIR}
+    LIBRARY  DESTINATION ${CMAKE_INSTALL_LIBDIR}
+    RUNTIME  DESTINATION ${CMAKE_INSTALL_BINDIR})
+install(DIRECTORY include/ DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
+
+install(EXPORT wolffConfig DESTINATION share/wolff/cmake)
+
+export(TARGETS wolff FILE wolffConfig.cmake)
diff --git a/lib/angle.h b/lib/angle.h
deleted file mode 100644
index c3f128e..0000000
--- a/lib/angle.h
+++ /dev/null
@@ -1,48 +0,0 @@
-#pragma once
-
-#include "types.h"
-
-#include <cmath>
-#include "vector.h"
-
-class angle_t {
-  public:
-    double x;
-
-    typedef vector_t<2, double> M_t;
-    typedef vector_t<2, double> F_t;
-
-    angle_t() : x(0) {}
-    angle_t(double x) : x(x) {}
-
-    inline vector_t<2, double> operator*(v_t a) const {
-      vector_t<2, double>M;
-      M[0] = a * cos(x);
-      M[1] = a * sin(x);
-
-      return M;
-    }
-
-    inline vector_t<2, double> operator*(double a) const {
-      vector_t<2, double>M;
-      M[0] = a * cos(x);
-      M[1] = a * sin(x);
-
-      return M;
-    }
-};
-
-inline vector_t<2,double>& operator+=(vector_t<2,double>& M, const angle_t& theta) {
-  M[0] += cos(theta.x);
-  M[1] += sin(theta.x);
-
-  return M;
-}
-
-inline vector_t<2,double>& operator-=(vector_t<2,double>& M, const angle_t& theta) {
-  M[0] -= cos(theta.x);
-  M[1] -= sin(theta.x);
-
-  return M;
-}
-
diff --git a/lib/circle_group.h b/lib/circle_group.h
deleted file mode 100644
index cb7cadd..0000000
--- a/lib/circle_group.h
+++ /dev/null
@@ -1,46 +0,0 @@
-#pragma once
-
-#include "angle.h"
-
-class circle_group_t {
-  public:
-    bool is_reflection;
-    double x;
-
-    circle_group_t() : is_reflection(false), x(0) {}
-    circle_group_t(bool x, double y) : is_reflection(x), x(y) {}
-
-    angle_t act(const angle_t& theta) const {
-      if (is_reflection) {
-        return angle_t(fmod(2 * M_PI + x - theta.x, 2 * M_PI));
-      } else {
-        return angle_t(fmod(x + theta.x, 2 * M_PI));
-      }
-    }
-
-    circle_group_t act(const circle_group_t& r) const {
-      if (is_reflection) {
-        return circle_group_t(!r.is_reflection, fmod(2 * M_PI + x - r.x, 2 * M_PI));
-      } else {
-        return circle_group_t(r.is_reflection, fmod(x + r.x, 2 * M_PI));
-      }
-    }
-
-    angle_t act_inverse(const angle_t& theta) const {
-      if (is_reflection) {
-        return act(theta);
-      } else {
-        return angle_t(fmod(2 * M_PI + theta.x - x, 2 * M_PI));
-      }
-    }
-
-    circle_group_t act_inverse(const circle_group_t& r) const {
-      if (is_reflection) {
-        return act(r);
-      } else {
-        return circle_group_t(r.is_reflection, fmod(2 * M_PI + r.x - x, 2 * M_PI));
-      }
-    }
-};
-
-
diff --git a/lib/convex.c b/lib/convex.c
deleted file mode 100644
index 7255ad4..0000000
--- a/lib/convex.c
+++ /dev/null
@@ -1,102 +0,0 @@
-
-#include "convex.h"
-
-double slope(point_t *P, point_t *Q) {
-  return (Q->y - P->y) / ((double)(Q->x) - (double)(P->x));
-}
-
-double *get_convex_minorant(count_t n, double *Gammas) {
-  if (n < 2) {
-    return Gammas;
-  }
-
-  list_t *L = (list_t *)calloc(1, sizeof(list_t));
-  L->p = (point_t *)calloc(1, sizeof(point_t));
-  L->p->x = 0;
-  L->p->y = Gammas[0];
-
-  list_t *pos = L;
-
-  for (count_t i = 1; i < n; i++) {
-    pos->next = (list_t *)calloc(1, sizeof(list_t));
-    pos->next->p = (point_t *)calloc(1, sizeof(point_t));
-    pos->next->p->x = i;
-    pos->next->p->y = Gammas[i];
-    pos->next->prev = pos;
-    pos = pos->next;
-  }
-
-  pos->next = (list_t *)calloc(1, sizeof(list_t));
-  pos->next->p = (point_t *)calloc(1, sizeof(point_t));
-  pos->next->p->x = n;
-  pos->next->p->y = 0;
-  pos->next->prev = pos;
-
-  list_t *X = L;
-  list_t *Y = L->next;
-  list_t *Z = Y->next;
-
-  while (true) {
-    if (slope(X->p, Y->p) <= slope(Y->p, Z->p)) {
-      X = Y;
-      Y = Z;
-      if (Z->next == NULL) {
-        break;
-      } else {
-        Z = Z->next;
-      }
-    } else {
-      Y->prev->next = Y->next;
-      Y->next->prev = Y->prev;
-      free(Y->p);
-      free(Y);
-      if (X->prev != NULL) {
-        Y = X;
-        X = X->prev;
-      } else {
-        if (Z->next != NULL) {
-          Y = Z;
-          Z = Z->next;
-        } else {
-          break;
-        }
-      }
-    }
-  }
-
-  pos = L;
-
-  double *g = (double *)calloc(n + 1, sizeof(double));
-  double rho = 0;
-
-  for (count_t i = 0; i < n + 1; i++) {
-    if (i > pos->next->p->x) {
-      pos = pos->next;
-    }
-
-    g[i] = pos->p->y + ((double)i - (double)(pos->p->x)) *  (pos->next->p->y - pos->p->y) / ((double)(pos->next->p->x) - (double)(pos->p->x));
-
-    if (i <n) {
-      if (Gammas[i] - g[i] > rho) {
-        rho = Gammas[i] - g[i];
-      }
-    } else {
-      if (0 - g[i] > rho) {
-        rho = 0 - g[i];
-      }
-    }
-  }
-
-  for (count_t i = 0; i < n + 1; i++) {
-    g[i] += rho / 2;
-  }
-
-  while (L != NULL) {
-    free(L->p);
-    L = L->next;
-    free(L);
-  }
-
-  return g;
-}
-
diff --git a/lib/convex.h b/lib/convex.h
deleted file mode 100644
index 5a405d4..0000000
--- a/lib/convex.h
+++ /dev/null
@@ -1,23 +0,0 @@
-
-#pragma once
-
-#include <inttypes.h>
-#include <stdbool.h>
-#include <stdlib.h>
-#include <string.h>
-
-#include "types.h"
-
-typedef struct {
-  count_t x;
-  double y;
-} point_t;
-
-typedef struct list_tag {
-  struct list_tag *prev;
-  struct list_tag *next;
-  point_t *p;
-} list_t;
-
-double *get_convex_minorant(count_t n, double *Gammas);
-
diff --git a/lib/wolff.h b/lib/include/wolff.hpp
index 141a5b2..c10a211 100644
--- a/lib/wolff.h
+++ b/lib/include/wolff.hpp
@@ -1,9 +1,9 @@
 
-#include "cluster.h"
-#include "state.h"
+#include "wolff/cluster.hpp"
+#include "wolff/state.hpp"
 
 template <class R_t, class X_t>
-void wolff(count_t N, state_t <R_t, X_t>& s, std::function <R_t(gsl_rng *, X_t s0)> gen_R, std::function <void(const state_t <R_t, X_t>&)> measurements, gsl_rng *r, bool silent) {
+void wolff(count_t N, state_t <R_t, X_t>& s, std::function <R_t(std::mt19937&, X_t)> gen_R, std::function <void(const state_t <R_t, X_t>&)> measurements, std::mt19937& r, bool silent) {
 
 #ifdef FINITE_STATES
 #ifdef NOFIELD
@@ -13,11 +13,13 @@ void wolff(count_t N, state_t <R_t, X_t>& s, std::function <R_t(gsl_rng *, X_t s
 #endif
 #endif
 
+  std::uniform_int_distribution<v_t> dist(0, s.nv);
+
   if (!silent) printf("\n");
   for (count_t steps = 0; steps < N; steps++) {
     if (!silent) printf("\033[F\033[JWOLFF: step %" PRIu64 " / %" PRIu64 ": E = %.2f, S = %" PRIv "\n", steps, N, s.E, s.last_cluster_size);
 
-    v_t v0 = gsl_rng_uniform_int(r, s.nv);
+    v_t v0 = dist(r);
     R_t step = gen_R(r, s.spins[v0]);
     flip_cluster <R_t, X_t> (s, v0, step, r);
 
diff --git a/lib/cluster.h b/lib/include/wolff/cluster.hpp
index f57bb68..104f3c2 100644
--- a/lib/cluster.h
+++ b/lib/include/wolff/cluster.hpp
@@ -1,18 +1,18 @@
 
 #pragma once
 
-#include <gsl/gsl_randist.h>
-#include <gsl/gsl_rng.h>
+#include <random>
 #include <cmath>
 #include <vector>
 #include <stack>
 
 #include "types.h"
-#include "state.h"
-#include "graph.h"
+#include "state.hpp"
+#include "graph.hpp"
 
 template <class R_t, class X_t>
-void flip_cluster(state_t<R_t, X_t>& s, v_t v0, const R_t& r, gsl_rng *rand) {
+void flip_cluster(state_t<R_t, X_t>& s, v_t v0, const R_t& r, std::mt19937& rand) {
+  std::uniform_real_distribution<double> dist(0.0,1.0);
   v_t nv = 0;
 
   std::stack<v_t> stack;
@@ -89,7 +89,7 @@ void flip_cluster(state_t<R_t, X_t>& s, v_t v0, const R_t& r, gsl_rng *rand) {
         prob = 1.0 - exp(-dE / s.T);
 #endif
 
-        if (gsl_rng_uniform(rand) < prob) {
+        if (dist(rand) < prob) {
           stack.push(vn); // push the neighboring vertex to the stack
         }
       }
diff --git a/lib/finite_states.h b/lib/include/wolff/finite_states.hpp
index 426edad..426edad 100644
--- a/lib/finite_states.h
+++ b/lib/include/wolff/finite_states.hpp
diff --git a/lib/graph.h b/lib/include/wolff/graph.hpp
index de06924..165544a 100644
--- a/lib/graph.h
+++ b/lib/include/wolff/graph.hpp
@@ -17,6 +17,5 @@ class graph_t {
 
     graph_t(D_t D, L_t L);
     void add_ext();
-
 };
 
diff --git a/lib/state.h b/lib/include/wolff/state.hpp
index cad453c..1f5e359 100644
--- a/lib/state.h
+++ b/lib/include/wolff/state.hpp
@@ -5,7 +5,7 @@
 #include <vector>
 
 #include "types.h"
-#include "graph.h"
+#include "graph.hpp"
 
 template <class R_t, class X_t>
 class state_t {
diff --git a/lib/types.h b/lib/include/wolff/types.h
index ec9efaf..ec9efaf 100644
--- a/lib/types.h
+++ b/lib/include/wolff/types.h
diff --git a/lib/rand.c b/lib/rand.c
deleted file mode 100644
index 76f537d..0000000
--- a/lib/rand.c
+++ /dev/null
@@ -1,20 +0,0 @@
-
-#include <assert.h>
-#include <stdio.h>
-
-unsigned long int rand_seed() {
-  /*
-    returns a random unsigned long integer read from the standard unix
-    pseudorandom device /dev/urandom
-  */
-
-  FILE *f = fopen("/dev/urandom", "r");
-  assert(f != NULL);
-
-  unsigned long int seed;
-  fread(&seed, sizeof(unsigned long int), 1, f);
-
-  fclose(f);
-
-  return seed;
-}
diff --git a/lib/rand.h b/lib/rand.h
deleted file mode 100644
index 7bb5354..0000000
--- a/lib/rand.h
+++ /dev/null
@@ -1,16 +0,0 @@
-
-#pragma once
-
-#include <assert.h>
-#include <stdio.h>
-
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-unsigned long int rand_seed();
-
-#ifdef __cplusplus
-}
-#endif
-
diff --git a/lib/graph.cpp b/lib/src/graph.cpp
index ca251f3..4043413 100644
--- a/lib/graph.cpp
+++ b/lib/src/graph.cpp
@@ -1,5 +1,5 @@
 
-#include "graph.h"
+#include <wolff/graph.hpp>
 
 graph_t::graph_t(D_t D, L_t L) {
   nv = pow(L, D);
diff --git a/lib/torus.h b/lib/torus.h
deleted file mode 100644
index 2aead52..0000000
--- a/lib/torus.h
+++ /dev/null
@@ -1,64 +0,0 @@
-
-#pragma once
-
-#include <cmath>
-#include <array>
-#include "types.h"
-
-template <q_t n>
-class torus_t : public std::array<double, n> {
-  public:
-    typedef std::array<double, n> M_t;
-    typedef std::array<double, n> F_t;
-
-    torus_t() {
-      this->fill(0);
-    }
-
-    inline torus_t<n> operator*(v_t a) const {
-      torus_t<n> x;
-      for (q_t i = 0; i < n; i++) {
-        x[i] = a * (*this)[i];
-      }
-
-      return x;
-    }
-
-    inline torus_t<n> operator*(double a) const {
-      torus_t<n> x;
-      for (q_t i = 0; i < n; i++) {
-        x[i] = a * (*this)[i];
-      }
-
-      return x;
-    }
-
-    inline torus_t<n>& operator+=(const torus_t<n>& x) {
-      for (q_t i = 0; i < n; i++) {
-        (*this)[i] += x[i];
-      }
-    }
-
-    inline torus_t<n>& operator-=(const torus_t<n>& x) {
-      for (q_t i = 0; i < n; i++) {
-        (*this)[i] -= x[i];
-      }
-    }
-};
-
-template <q_t n>
-double norm_squared(const torus_t<n>& x) {
-  double tmp = 0;
-  for (const double& xi : x) {
-    tmp += pow(xi, 2);
-  }
-  return tmp;
-}
-
-void write_magnetization(const torus_t<n>& x, FILE *outfile) {
-  for (const double& xi : x) {
-    float tmp_xi = (float)xi;
-    fwrite(&tmp_xi, sizeof(float), 1, outfile);
-  }
-}
-
diff --git a/wolfram_link/Makefile b/wolfram_link/Makefile
deleted file mode 100644
index 9d65623..0000000
--- a/wolfram_link/Makefile
+++ /dev/null
@@ -1,13 +0,0 @@
-
-CC = clang
-WSPREP = /opt/Mathematica/SystemFiles/Links/WSTP/DeveloperKit/Linux-x86-64/CompilerAdditions/wsprep
-CFLAGS = -g -Os -O3 -Wall -fno-strict-aliasing -Wstrict-overflow -Wno-missing-field-initializers -flto -fopenmp=libiomp5 -I/usr/lib/gcc/x86_64-linux-gnu/4.8/include/ -march=native -I/opt/Mathematica/SystemFiles/Links/WSTP/DeveloperKit/Linux-x86-64/CompilerAdditions/ -I../lib/
-LDFLAGS = -L/opt/Mathematica/SystemFiles/Links/WSTP/DeveloperKit/Linux-x86-64/CompilerAdditions -lm -lpthread -lrt -lstdc++ -ldl -luuid -l WSTP64i4
-
-convex.o: ../lib/convex.c
-	@${CC} -c -o $@ $< ${CFLAGS}
-
-convexminorant: convexminorant.tm convex.o
-	@${WSPREP} $< -o $@.c
-	@${CC} -o $@ convex.o $@.c ${CFLAGS} ${LDFLAGS} 
-
diff --git a/wolfram_link/convexminorant.tm b/wolfram_link/convexminorant.tm
deleted file mode 100644
index edca6f4..0000000
--- a/wolfram_link/convexminorant.tm
+++ /dev/null
@@ -1,34 +0,0 @@
-
-#include <convex.h>
-#include <wstp.h>
-
-extern int WSMain(int, char **);
-
-void convexminorant(double *, int);
-
-:Begin:
-:Function:      convexminorant
-:Pattern:       GetConvexMinorant[ list:{___Real} ]
-:Arguments:     { list }
-:ArgumentTypes: { Real64List }
-:ReturnType:    Manual
-:End:
-
-:Evaluate:      GetConvexMinorant[ sequence___Float]:= GetConvexMinorant[ {sequence} ]
-
-void convexminorant(double * Gammas, int len) {
-  int i;
-  for (i = 0; i < len; i++) {
-    if (Gammas[i] <= 0) {
-       break;
-    }
-  }
-  double *m = get_convex_minorant(i, Gammas);
-  WSPutReal64List(stdlink, m, i);
-  free(m);
-}
-
-int main(int argc, char **argv) {
-  return WSMain(argc, argv);
-}
-