chore: replace postfix increment and decrement operators with their p…

…refix equivalents

cpp/examples/forward_backward/inpainting.cc

@@ -80,7 +80,7 @@ int main(int argc, char const **argv) {
   SOPT_HIGH_LOG("Create dirty vector");
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(mersenne);
   // Write dirty imagte to file
   if (output != "none") {
     Vector const dirty = sampling.adjoint() * y;

cpp/examples/forward_backward/inpainting_credible_interval.cc

@@ -82,7 +82,7 @@ int main(int argc, char const **argv) {
   SOPT_HIGH_LOG("Create dirty vector");
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(mersenne);
   // Write dirty imagte to file
   if (output != "none") {
     Vector const dirty = sampling.adjoint() * y;

cpp/examples/forward_backward/inpainting_joint_map.cc

@@ -81,7 +81,7 @@ int main(int argc, char const **argv) {
   SOPT_HIGH_LOG("Create dirty vector");
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(mersenne);
   // Write dirty imagte to file
   if (output != "none") {
     Vector const dirty = sampling.adjoint() * y;

cpp/examples/forward_backward/l2_inpainting.cc

@@ -78,7 +78,7 @@ int main(int argc, char const **argv) {
   SOPT_HIGH_LOG("Create dirty vector");
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(mersenne);
   // Write dirty imagte to file
   if (output != "none") {
     Vector const dirty = sampling.adjoint() * y;

cpp/examples/primal_dual/inpainting.cc

@@ -89,7 +89,7 @@ int main(int argc, char const **argv) {
   SOPT_HIGH_LOG("Create dirty vector");
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(mersenne);
   // Write dirty image to file
   if (output != "none") {
     Vector const dirty = sampling.adjoint() * y;

cpp/examples/primal_dual/tv_inpainting.cc

@@ -82,7 +82,7 @@ int main(int argc, char const **argv) {
   SOPT_HIGH_LOG("Create dirty vector");
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(mersenne);
   // Write dirty image to file
   if (output != "none") {
     Vector const dirty = sampling.adjoint() * y;

cpp/examples/proximal_admm/inpainting.cc

@@ -75,7 +75,7 @@ int main(int argc, char const **argv) {
   SOPT_HIGH_LOG("Create dirty vector");
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(mersenne);
   // Write dirty imagte to file
   if (output != "none") {
     Vector const dirty = sampling.adjoint() * y;

cpp/examples/proximal_admm/reweighted.cc

@@ -79,7 +79,7 @@ int main(int argc, char const **argv) {
   SOPT_MEDIUM_LOG("Create dirty vector");
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(mersenne);
   // Write dirty imagte to file
   if (output != "none") {
     Vector const dirty = sampling.adjoint() * y;

cpp/examples/sdmm/inpainting.cc

@@ -74,7 +74,7 @@ int main(int argc, char const **argv) {
   SOPT_HIGH_LOG("Create dirty vector");
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(mersenne);
   // Write dirty imagte to file
   if (output != "none") {
     Vector const dirty = sampling.adjoint() * y;

cpp/examples/sdmm/reweighted.cc

@@ -78,7 +78,7 @@ int main(int argc, char const **argv) {
   SOPT_HIGH_LOG("Create dirty vector");
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(mersenne);
   // Write dirty imagte to file
   if (output != "none") {
     Vector const dirty = sampling.adjoint() * y;

cpp/sopt/chained_operators.h

@@ -23,12 +23,12 @@ OperatorFunction<T0> chained_operators(OperatorFunction<T0> const &arg0, T const
       (*first)(output, input);
     else {
       (*first)(*buffer, input);
-      first++;
+      ++first;
       (*first)(output, *buffer);
     }
-    for (++first; first != last; first++) {
+    for (++first; first != last; ++first) {
       (*first)(*buffer, output);
-      first++;
+      ++first;
       (*first)(output, *buffer);
     }
   };

cpp/sopt/cppflow_utils.cc

@@ -70,7 +70,7 @@ namespace cppflowutils {
   cppflow::tensor convert_image_to_tensor(sopt::Vector<std::complex<double>> const &image, int image_rows, int image_cols) {
 
     std::vector<float> input_values(image_rows);
-    for(int i = 0; i < image_rows; i++)
+    for(int i = 0; i < image_rows; ++i)
     {
       if(std::abs(image(i).real()) > cppflowutils::imaginary_threshold)
       {

cpp/sopt/credible_region.h

@@ -145,8 +145,8 @@ credible_interval_grid(const Eigen::MatrixBase<T> &solution, const t_uint &rows,
   Image<K> credible_grid_upper_bound = Image<K>::Zero(grid_rows, grid_cols);
   Image<K> credible_grid_mean = Image<K>::Zero(grid_rows, grid_cols);
   SOPT_LOW_LOG("Starting calculation of credible interval: {} x {} grid.", grid_rows, grid_cols);
-  for (t_uint i = 0; i < grid_rows; i++) {
-    for (t_uint j = 0; j < grid_cols; j++) {
+  for (t_uint i = 0; i < grid_rows; ++i) {
+    for (t_uint j = 0; j < grid_cols; ++j) {
       const t_uint start_row = i * drow;
       const t_uint start_col = j * dcol;
       if (static_cast<t_int>(rows - start_row - drow) < 0)

cpp/sopt/gradient_operator.h

@@ -28,9 +28,9 @@ template <class T>
 Vector<T> diff2d(const Vector<T> &x, const t_int rows, const t_int cols) {
   Matrix<T> output = Matrix<T>::Zero(rows, 2 * cols);
   const Matrix<T> &input_image = Matrix<T>::Map(x.data(), rows, cols);
-  for (Eigen::Index i(0); i < rows; i++)
+  for (Eigen::Index i(0); i < rows; ++i)
     output.block(0, 0, rows, cols).row(i) = diff<T>(input_image.row(i));
-  for (Eigen::Index i(0); i < cols; i++)
+  for (Eigen::Index i(0); i < cols; ++i)
     output.block(0, cols, rows, cols).col(i) = diff<T>(input_image.col(i));
   return Vector<T>::Map(output.data(), output.size());
 }
@@ -39,9 +39,9 @@ template <class T>
 Vector<T> diff2d_adjoint(const Vector<T> &x, const t_int rows, const t_int cols) {
   const Matrix<T> &input_image = Matrix<T>::Map(x.data(), rows, 2 * cols);
   Matrix<T> output = Matrix<T>::Zero(rows, cols);
-  for (Eigen::Index i(0); i < rows; i++)
+  for (Eigen::Index i(0); i < rows; ++i)
     output.row(i) += diff_adjoint<T>(input_image.block(0, 0, rows, cols).row(i));
-  for (Eigen::Index i(0); i < cols; i++)
+  for (Eigen::Index i(0); i < cols; ++i)
     output.col(i) += diff_adjoint<T>(input_image.block(0, cols, rows, cols).col(i));
   return Vector<T>::Map(output.data(), output.size());
 }

cpp/sopt/imaging_primal_dual.h

@@ -318,7 +318,7 @@ typename ImagingPrimalDual<SCALAR>::Diagnostic ImagingPrimalDual<SCALAR>::operat
   auto const g_proximal = [this](t_Vector &out, Real gamma, t_Vector const &x) {
     this->l2ball_proximal()(out, gamma, x);
     // applying preconditioning
-    for (t_int i = 0; i < this->precondition_iters(); i++)
+    for (t_int i = 0; i < this->precondition_iters(); ++i)
       this->l2ball_proximal()(
           out, gamma,
           out - this->precondition_stepsize() *

cpp/sopt/joint_map.h

@@ -104,7 +104,7 @@ class JointMAP {
     typedef typename ALGORITHM::DiagnosticAndResult ResultType;
     ResultType result = (*(this->algo_ptr_))(std::forward<ARGS>(args)...);
     t_real gamma = 0;
-    niters++;
+    ++niters;
     t_uint algo_iters(result.niters);
     for (; (not converged) && (niters < itermax()); ++niters) {
       SOPT_LOW_LOG("    - [JMAP] Iteration {}/{}", niters, itermax());

cpp/sopt/l2_primal_dual.h

@@ -317,7 +317,7 @@ typename ImagingPrimalDual<SCALAR>::Diagnostic ImagingPrimalDual<SCALAR>::operat
   auto const g_proximal = [this](t_Vector &out, Real gamma, t_Vector const &x) {
     this->l2ball_proximal()(out, gamma, x);
     // applying preconditioning
-    for (t_int i = 0; i < this->precondition_iters(); i++)
+    for (t_int i = 0; i < this->precondition_iters(); ++i)
       this->l2ball_proximal()(
           out, gamma,
           out - this->precondition_stepsize() *

cpp/sopt/mpi/communicator.h

@@ -334,7 +334,7 @@ Communicator::all_to_allv(const std::vector<T> &vec, std::vector<t_int> const &s
     return vec;
   }
   std::vector<t_int> rec_sizes(send_sizes.size(), 0);
-  for (t_int i = 0; i < size(); i++) {
+  for (t_int i = 0; i < size(); ++i) {
     if (i == rank())
       rec_sizes = gather<t_int>(send_sizes[i], i);
     else
@@ -383,7 +383,7 @@ typename std::enable_if<is_registered_type<T>::value, Vector<T>>::type Communica
     return vec;
   }
   std::vector<t_int> rec_sizes(send_sizes.size(), 0);
-  for (t_int i = 0; i < size(); i++) {
+  for (t_int i = 0; i < size(); ++i) {
     if (i == rank())
       rec_sizes = gather<t_int>(send_sizes[i], i);
     else

cpp/sopt/proximal.h

@@ -101,7 +101,7 @@ void tv_norm(Eigen::DenseBase<T0> &out, typename real_type<typename T0::Scalar>:
   typename T1::PlainObject const &u = x.segment(0, size);
   typename T1::PlainObject const &v = x.segment(size, size);
   out = T0::Zero(size * 2);
-  for (typename Eigen::Index i(0); i < size; i++) {
+  for (typename Eigen::Index i(0); i < size; ++i) {
     const t_real norm = std::sqrt(std::abs(u(i) * u(i) + v(i) * v(i)));
     if (norm > gamma) {
       out(i) = (1 - gamma / norm) * u(i);
@@ -119,7 +119,7 @@ void tv_norm(Eigen::DenseBase<T0> &out, Eigen::DenseBase<T2> const &gamma,
   typename T1::PlainObject const &u = x.segment(0, size);
   typename T1::PlainObject const &v = x.segment(size, size);
   out = T0::Zero(size * 2);
-  for (typename Eigen::Index i(0); i < size; i++) {
+  for (typename Eigen::Index i(0); i < size; ++i) {
     const t_real norm = std::sqrt(std::abs(u(i) * u(i) + v(i) * v(i)));
     if (norm > gamma(i)) {
       out(i) = (1 - gamma(i) / norm) * u(i);

cpp/sopt/sdmm.h

@@ -275,7 +275,7 @@ void SDMM<SCALAR>::update_directions(t_Vectors &y, t_Vectors &z, t_Vector const
 template <class SCALAR>
 void SDMM<SCALAR>::initialization(t_Vectors &y, t_Vectors &z, t_Vector const &x) const {
   SOPT_TRACE("Initializing SDMM");
-  for (t_uint i(0); i < transforms().size(); i++) {
+  for (t_uint i(0); i < transforms().size(); ++i) {
     y[i] = transforms(i) * x;
     z[i].resize(y[i].size());
     z[i].fill(0);

cpp/sopt/tf_g_proximal.h

@@ -106,7 +106,7 @@ class TFGProximal : public GProximal<SCALAR> {
     // https://stackoverflow.com/questions/3786360/confusing-template-error
     auto output_vector = model_output[0].template get_data<float>();
 
-    for(int i = 0; i < image_size; i++) {
+    for(int i = 0; i < image_size; ++i) {
       image_out[i] = static_cast<Scalar>(output_vector[i]);
     }
   }

cpp/sopt/tv_primal_dual.h

@@ -316,7 +316,7 @@ typename TVPrimalDual<SCALAR>::Diagnostic TVPrimalDual<SCALAR>::operator()(
   auto const g_proximal = [this](t_Vector &out, Real gamma, t_Vector const &x) {
     this->l2ball_proximal()(out, gamma, x);
     // applying preconditioning
-    for (t_int i = 0; i < this->precondition_iters(); i++)
+    for (t_int i = 0; i < this->precondition_iters(); ++i)
       this->l2ball_proximal()(
           out, gamma,
           out - this->precondition_stepsize() *

cpp/sopt/wavelets/innards.impl.h

@@ -128,7 +128,7 @@ void up_convolve_sum(Eigen::ArrayBase<T0> &result, Eigen::ArrayBase<T1> const &c
 #ifdef SOPT_OPENMP
 #pragma omp parallel for
 #endif
-  for (typename T0::Index i = 0; i < result.size() / 2; i++) {
+  for (typename T0::Index i = 0; i < result.size() / 2; ++i) {
     result(2 * i + index_place_even) =
         periodic_scalar_product(coeffs.head(Nlow), low_even, i + even_offset) +
         periodic_scalar_product(coeffs.tail(Nhigh), high_even, i + even_offset);

cpp/tests/communicator.cc

@@ -60,7 +60,7 @@ TEST_CASE("Creates an mpi communicator") {
       auto const result = world.scatter_one(scattered);
       REQUIRE(result == world.rank() + 2);
       auto const gathered = world.gather(result);
-      for (decltype(gathered)::size_type i = 0; i < gathered.size(); i++)
+      for (decltype(gathered)::size_type i = 0; i < gathered.size(); ++i)
         CHECK(gathered[i] == scattered[i]);
     } else {
       auto const result = world.scatter_one<t_int>();
@@ -76,7 +76,7 @@ TEST_CASE("Creates an mpi communicator") {
     Vector<t_int> const sendee = Vector<t_int>::Constant(size(world.rank()), world.rank());
     std::vector<t_int> sizes(world.size());
     int n(0);
-    std::generate(sizes.begin(), sizes.end(), [&n, &size]() { return size(n++); });
+    std::generate(sizes.begin(), sizes.end(), [&n, &size]() { return size(++n); });
 
     auto const result = world.is_root() ? world.gather(sendee, sizes) : world.gather(sendee);
     if (world.rank() == world.root_id()) {
@@ -160,7 +160,7 @@ TEST_CASE("Creates an mpi communicator") {
           Vector<t_int>::Constant(std::accumulate(sizes.begin(), sizes.end(), 0), world.rank());
       const Vector<t_int> output = world.all_to_allv(sendee, sizes);
       t_int sum = 0;
-      for (t_int i = 0; i < world.size() - 1; i++) {
+      for (t_int i = 0; i < world.size() - 1; ++i) {
         const Vector<t_int> expected = Vector<t_int>::Constant(i + 1, i + 1);
         CAPTURE(sum);
         CAPTURE(i);

cpp/tests/credible_region.cc

@@ -19,7 +19,7 @@ TEST_CASE("calculating gamma") {
   };
   const t_Vector x = t_Vector::Random(N);
   CHECK(0 == energy_function(x));
-  for (t_uint i = 1; i < 10; i++) {
+  for (t_uint i = 1; i < 10; ++i) {
     const t_real alpha = 0.9 + i * 0.01;
     const t_real gamma = credible_region::compute_energy_upper_bound(alpha, x, energy_function);
     CHECK(gamma == Approx(N * (std::sqrt(16 * std::log(3 / (1 - alpha)) / N) + 1)));

cpp/tests/gradient_operator.cc

@@ -26,7 +26,7 @@ TEST_CASE("Gradient Operator") {
   Image const image = sopt::notinstalled::read_standard_tiff("cameraman256");
   auto const psi = sopt::gradient_operator::gradient_operator<Scalar>(image.rows(), image.cols());
   Matrix input = Matrix::Ones(image.rows(), image.cols());
-  for (Eigen::Index i(0); i < image.rows(); i++) input.row(i) *= static_cast<Scalar>(i);
+  for (Eigen::Index i(0); i < image.rows(); ++i) input.row(i) *= static_cast<Scalar>(i);
   Vector output = psi.adjoint() * Vector::Map(input.data(), input.size());
   CAPTURE(output.segment(0, 5));
   CAPTURE(output.segment(image.size(), 5));
@@ -35,7 +35,7 @@ TEST_CASE("Gradient Operator") {
   CHECK(output.segment(input.size(), input.size() - 1)
             .isApprox(Vector::Constant(0.5, input.size() - 1)));
   input = Matrix::Ones(image.rows(), image.cols());
-  for (Eigen::Index i(0); i < image.cols(); i++) input.col(i) *= static_cast<Scalar>(i);
+  for (Eigen::Index i(0); i < image.cols(); ++i) input.col(i) *= static_cast<Scalar>(i);
   output = psi.adjoint() * Vector::Map(input.data(), input.size());
   CAPTURE(output.segment(0, 5));
   CAPTURE(output.segment(image.size(), 5));

cpp/tests/inpainting.cc

@@ -50,7 +50,7 @@ TEST_CASE("Inpainting"){
 
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(*mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(*mersenne);
 
   sopt::t_real const gamma = 18;
   sopt::t_real const beta = sigma * sigma * 0.5;

cpp/tests/serial_vs_parallel_padmm.cc

@@ -69,7 +69,7 @@ TEST_CASE("Parallel vs serial inpainting") {
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y = world.is_root() ? y0 : world.broadcast<Vector>();
   if (world.is_root()) {
-    for (sopt::t_int i = 0; i < y0.size(); i++) y(i) += gaussian_dist(*mersenne);
+    for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) += gaussian_dist(*mersenne);
     world.broadcast(y);
   }
   if (split_comm.size() > 1) {

cpp/tests/tf_inpainting.cc

@@ -46,7 +46,7 @@ TEST_CASE("Inpainting"){
 
   std::normal_distribution<> gaussian_dist(0, sigma);
   Vector y(y0.size());
-  for (sopt::t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(*mersenne);
+  for (sopt::t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(*mersenne);
 
   sopt::t_real const gamma = 18;
   sopt::t_real const beta = sigma * sigma * 0.5;

cpp/tests/wavelets.cc

@@ -60,7 +60,7 @@ void check_round_trip(Eigen::ArrayBase<T0> const &input_, sopt::t_uint db,
 }
 
 TEST_CASE("wavelet data") {
-  for (sopt::t_int num = 1; num < 100; num++) {
+  for (sopt::t_int num = 1; num < 100; ++num) {
     if (num < 39)
       REQUIRE(sopt::wavelets::daubechies_data(num).coefficients.size() == 2 * num);
     else

cpp/tools_for_tests/inpainting.h

@@ -29,7 +29,7 @@ Vector<T> dirty(sopt::LinearTransform<Vector<T>> const &sampling, sopt::Image<T>
   auto const y0 = target(sampling, image);
   std::normal_distribution<> gaussian_dist(0, sigma(sampling, image));
   Vector<T> y(y0.size());
-  for (t_int i = 0; i < y0.size(); i++) y(i) = y0(i) + gaussian_dist(mersenne);
+  for (t_int i = 0; i < y0.size(); ++i) y(i) = y0(i) + gaussian_dist(mersenne);
 
   return y;
 }