Draft MPI version

2021-08-19 14:48:10 +03:00
30 changed files with 533 additions and 4430 deletions
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@ -10,6 +10,8 @@ include(FindPkgConfig)
 include(FindPackageHandleStandardArgs)
 include(color_msg)

+find_package(MPI)
+
 option(BUILD_SHARED_LIBS "Build shared libraries." OFF)
 option(BUILD_STATIC_LIBS "Build static libraries." ON)
 option(ENABLE_SHARP "Enable SHARP support." ON)
@ -42,17 +44,6 @@ ENDIF(BUILD_PYTHON)

 MESSAGE(STATUS "Using the target ${CosmoTool_local} to build python module")

-find_library(ZLIB_LIBRARY z)
-find_library(DL_LIBRARY dl)
-find_library(RT_LIBRARY rt)
-if (RT_LIBRARY)
-  SET(RT_DEP rt)
-  message(STATUS "RT found, linking against it")
-else()
-  SET(RT_DEP)
-endif()
-find_library(MATH_LIBRARY m)
-
 include(${CMAKE_SOURCE_DIR}/external/external_build.cmake)

 IF(YORICK_SUPPORT)
@ -66,6 +57,9 @@ IF(YORICK_SUPPORT)
 ENDIF(YORICK_SUPPORT)

 find_program(CYTHON NAMES cython3 cython)
+find_library(ZLIB_LIBRARY z)
+find_library(DL_LIBRARY dl)
+find_library(MATH_LIBRARY m)

 set(NETCDF_FIND_REQUIRED ${YORICK_SUPPORT})
 set(GSL_FIND_REQUIRED TRUE)
@ -78,12 +72,14 @@ SET(CPACK_PACKAGE_DESCRIPTION_SUMMARY "A toolbox for impatient cosmologists")
 SET(CPACK_PACKAGE_VENDOR "Guilhem Lavaux")
 SET(CPACK_RESOURCE_FILE_LICENSE "${CMAKE_CURRENT_SOURCE_DIR}/LICENCE_CeCILL_V2")
 SET(CPACK_PACKAGE_VERSION_MAJOR "1")
-SET(CPACK_PACKAGE_VERSION_MINOR "3")
-SET(CPACK_PACKAGE_VERSION_PATCH "4${EXTRA_VERSION}")
+SET(CPACK_PACKAGE_VERSION_MINOR "2")
+SET(CPACK_PACKAGE_VERSION_PATCH "3${EXTRA_VERSION}")
 SET(CPACK_PACKAGE_INSTALL_DIRECTORY "CosmoToolbox-${CPACK_PACKAGE_VERSION_MAJOR}.${CPACK_PACKAGE_VERSION_MINOR}")
 SET(CPACK_STRIP_FILES "lib/libCosmoTool.so")
-SET(CPACK_SOURCE_IGNORE_FILES 
-"/CVS/;/\\\\.git/;/\\\\.svn/;\\\\.swp$;\\\\.#;/#;.*~;cscope.*;/CMakeFiles/;.*\\\\.cmake;Makefile")
+SET(CPACK_SOURCE_IGNORE_FILES "/CVS/;/\\\\.git/;/\\\\.svn/;\\\\.swp$;\\\\.#;/#;.*~;cscope.*;/CMakeFiles/;.*\\\\.cmake;Makefile")
+
+include_directories( ${MPI_C_INCLUDE_PATH})
+

 add_subdirectory(src)
 add_subdirectory(sample)
--- a/MANIFEST.in
+++ b/MANIFEST.in
@ -1,5 +1,4 @@
 include  .gitignore
-include  pyproject.toml
 include  CMakeLists.txt
 include  FindNumPy.cmake
 include  FindPyLibs.cmake
@ -17,17 +16,14 @@ include  doc/source/cpplibrary.rst
 include  doc/source/index.rst
 include  doc/source/intro.rst
 include  doc/source/pythonmodule.rst
-include  external/config.guess
-include  external/config.sub
 include  external/external_build.cmake
-include  external/libsharp-8d51946.tar.gz
+include  external/libsharp-6077806.tar.gz
 include  external/omptl-20120422.tar.bz2
 include  external/patch-omptl
 include  python/CMakeLists.txt
 include  python/_cosmo_bispectrum.cpp
 include  python/_cosmo_cic.pyx
 include  python/_cosmo_power.pyx
-include  python/_cosmomath.pyx
 include  python/_cosmotool.pyx
 include  python/_fast_interp.pyx
 include  python/_project.pyx
@ -97,7 +93,6 @@ include  sample/testkd3.cpp
 include  setup.py
 include  src/CMakeLists.txt
 include  src/algo.hpp
-include  src/numpy_adaptors.hpp
 include  src/bqueue.hpp
 include  src/bqueue.tcc
 include  src/bsp_simple.hpp
--- a/builder/build-wheels.sh
+++ b/builder/build-wheels.sh
@ -8,15 +8,12 @@ export CC CXX

 # Install a system package required by our library
 #yum install -y atlas-devel
-yum install -y cmake3 gsl-devel zlib-devel fftw3-devel libffi-devel hdf5 hdf5-devel
+yum install -y cmake3 gsl-devel zlib-devel fftw3-devel

 ln -fs /usr/bin/cmake3 /usr/bin/cmake


-test -d /io/wheelhouse || mkdir /io/wheelhouse
-test -d /io/wheelhouse/fix || mkdir /io/wheelhouse/fix
-
-ALL_PYTHON="cp39-cp39 cp310-cp310"
+ALL_PYTHON="cp36-cp36m cp37-cp37m cp38-cp38 cp39-cp39" 

 # Compile wheels
 for pkg in $ALL_PYTHON; do
@ -24,12 +21,12 @@ for pkg in $ALL_PYTHON; do
 #    "${PYBIN}/pip" install -r /io/dev-requirements.txt
    "${PYBIN}/pip" install setuptools wheel Cython 
    "${PYBIN}/pip" install -r /io/requirements.txt
-    "${PYBIN}/pip" wheel -vvv /io/ -w /io/wheelhouse/
+    "${PYBIN}/pip" wheel -vvv /io/ -w wheelhouse/
 done

 # Bundle external shared libraries into the wheels
-for whl in /io/wheelhouse/cosmotool*linux*.whl; do
-    auditwheel repair "$whl" --plat $PLAT -w /io/wheelhouse/fix
+for whl in wheelhouse/cosmotool*.whl; do
+    auditwheel repair "$whl" --plat $PLAT -w /io/wheelhouse/
 done

 # Install packages and test
--- a/builder/start.sh
+++ b/builder/start.sh
@ -9,4 +9,4 @@ if ! [ -e ${d}/setup.py ] ; then
  exit 1
 fi

-podman run -ti  --rm -e PLAT=manylinux2014_x86_64 -v ${d}:/io:Z quay.io/pypa/manylinux2014_x86_64 /io/builder/build-wheels.sh
+podman run -ti  --rm -e PLAT=manylinux2010_x86_64 -v ${d}:/io:Z quay.io/pypa/manylinux2010_x86_64 /io/builder/build-wheels.sh
--- a/conda/meta.yaml
+++ b/conda/meta.yaml
@ -1,6 +1,6 @@
 package:
  name: cosmotool
-  version: "1.3.0"
+  version: "1.2.3"

 source:
  git_rev: a86c9a8
--- a/external/config.guess
+++ b/external/config.guess
--- a/external/config.sub
+++ b/external/config.sub
--- a/external/external_build.cmake
+++ b/external/external_build.cmake
@ -2,19 +2,16 @@ include(FindOpenMP)

 OPTION(ENABLE_OPENMP "Set to Yes if Healpix and/or you need openMP" OFF)

-SET(SOURCE_PREFIX ${CMAKE_SOURCE_DIR})
 SET(FFTW_URL "http://www.fftw.org/fftw-3.3.3.tar.gz" CACHE STRING "URL to download FFTW from")
 SET(EIGEN_URL "https://gitlab.com/libeigen/eigen/-/archive/3.3.7/eigen-3.3.7.tar.bz2" CACHE STRING "URL to download Eigen from")
 SET(GENGETOPT_URL "ftp://ftp.gnu.org/gnu/gengetopt/gengetopt-2.22.5.tar.gz" CACHE STRING "URL to download gengetopt from")
-SET(HDF5_URL "https://support.hdfgroup.org/ftp/HDF5/releases/hdf5-1.12/hdf5-1.12.2/src/hdf5-1.12.2.tar.gz" CACHE STRING "URL to download HDF5 from")
-SET(NETCDF_URL "https://downloads.unidata.ucar.edu/netcdf-c/4.9.2/netcdf-c-4.9.2.tar.gz" CACHE STRING "URL to download NetCDF from")
-SET(NETCDFCXX_URL "https://github.com/Unidata/netcdf-cxx4/archive/v4.3.1.tar.gz" CACHE STRING "URL to download NetCDF-C++ from")
-SET(BOOST_URL "https://boostorg.jfrog.io/artifactory/main/release/1.82.0/source/boost_1_82_0.tar.gz" CACHE STRING "URL to download Boost from")
-SET(GSL_URL "https://ftpmirror.gnu.org/gsl/gsl-2.7.tar.gz" CACHE STRING "URL to download GSL from ")
+SET(HDF5_URL "https://support.hdfgroup.org/ftp/HDF5/releases/hdf5-1.8/hdf5-1.8.18/src/hdf5-1.8.18.tar.bz2" CACHE STRING "URL to download HDF5 from")
+SET(NETCDF_URL "ftp://ftp.unidata.ucar.edu/pub/netcdf/netcdf-4.5.0.tar.gz" CACHE STRING "URL to download NetCDF from")
+SET(NETCDFCXX_URL "https://github.com/Unidata/netcdf-cxx4/archive/v4.3.0.tar.gz" CACHE STRING "URL to download NetCDF-C++ from")
+SET(BOOST_URL "https://boostorg.jfrog.io/artifactory/main/release/1.74.0/source/boost_1_74_0.tar.bz2" CACHE STRING "URL to download Boost from")
+SET(GSL_URL "ftp://ftp.gnu.org/gnu/gsl/gsl-1.15.tar.gz" CACHE STRING "URL to download GSL from ")
 mark_as_advanced(FFTW_URL EIGEN_URL HDF5_URL NETCDF_URL BOOST_URL GSL_URL)

-file(MAKE_DIRECTORY ${SOURCE_PREFIX}/downloads)
-
 SET(all_deps)

 MACRO(CHECK_CHANGE_STATE VAR) 
@ -39,13 +36,12 @@ CHECK_CHANGE_STATE(INTERNAL_DLIB DLIB_INCLUDE_DIR DLIB_LIBRARIES)

 IF(ENABLE_OPENMP)
  IF (NOT OPENMP_FOUND)
-    MESSAGE(NOTICE "No known compiler option for enabling OpenMP")
-  ELSE()
-    SET(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} ${OpenMP_C_FLAGS}")
-    SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} ${OpenMP_CXX_FLAGS}")
-    SET(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} ${OpenMP_C_FLAGS}")
+    MESSAGE(ERROR "No known compiler option for enabling OpenMP")
  ENDIF(NOT OPENMP_FOUND)

+  SET(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} ${OpenMP_C_FLAGS}")
+  SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} ${OpenMP_CXX_FLAGS}")
+  SET(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} ${OpenMP_C_FLAGS}")
 ENDIF(ENABLE_OPENMP)


@ -59,25 +55,17 @@ if (ENABLE_SHARP)
  SET(DEP_BUILD ${BUILD_PREFIX}/sharp-prefix/src/sharp/auto)
  IF(NOT ENABLE_OPENMP)
    SET(SHARP_OPENMP --disable-openmp)
-  ELSE()
-    SET(SHARP_OPENMP)
  ENDIF()
  SET(CUTILS_LIBRARY ${DEP_BUILD}/lib/libc_utils.a)
  SET(FFTPACK_LIBRARY ${DEP_BUILD}/lib/libfftpack.a)
  SET(SHARP_LIBRARY ${DEP_BUILD}/lib/libsharp.a)
  SET(SHARP_LIBRARIES ${SHARP_LIBRARY} ${FFTPACK_LIBRARY} ${CUTILS_LIBRARY})
  SET(SHARP_INCLUDE_PATH ${DEP_BUILD}/include)
-  message(STATUS "Flags: ${CMAKE_C_FLAGS}")
  ExternalProject_Add(sharp
-    URL ${CMAKE_SOURCE_DIR}/external/libsharp-8d51946.tar.gz
+    URL ${CMAKE_SOURCE_DIR}/external/libsharp-6077806.tar.gz
    PREFIX ${BUILD_PREFIX}/sharp-prefix
-    DOWNLOAD_DIR ${SOURCE_PREFIX}/downloads
    BUILD_IN_SOURCE 1 
-    CONFIGURE_COMMAND
-          cp -f ${CMAKE_SOURCE_DIR}/external/config.guess . &&
-          cp -f ${CMAKE_SOURCE_DIR}/external/config.sub . &&
-          autoconf &&
-          ./configure "CC=${CMAKE_C_COMPILER}" "CXX=${CMAKE_CXX_COMPILER}" "CFLAGS=${CMAKE_C_FLAGS}" "LDFLAGS=${CMAKE_EXE_LINKER_FLAGS}" --prefix=${DEP_BUILD} ${SHARP_OPENMP}
+    CONFIGURE_COMMAND autoconf && ./configure "CC=${CMAKE_C_COMPILER}" "CXX=${CMAKE_CXX_COMPILER}" --prefix=${DEP_BUILD} ${SHARP_OPENMP}
    INSTALL_COMMAND echo "No install"
    BUILD_BYPRODUCTS ${SHARP_LIBRARIES}
  )
@ -94,23 +82,27 @@ if (INTERNAL_HDF5)
  ExternalProject_Add(hdf5
    PREFIX ${BUILD_PREFIX}/hdf5-prefix
    URL ${HDF5_URL}
-    URL_HASH MD5=30172c75e436d7f2180e274071a4ca97
-    DOWNLOAD_DIR ${SOURCE_PREFIX}/downloads
-    CONFIGURE_COMMAND
-        ${HDF5_SOURCE_DIR}/configure "CC=${CMAKE_C_COMPILER}" "CXX=${CMAKE_CXX_COMPILER}" "CFLAGS=${CMAKE_C_FLAGS}" "LDFLAGS=${CMAKE_EXE_LINKER_FLAGS}" --prefix=${HDF5_BIN_DIR} --disable-shared  --enable-cxx --enable-hl --enable-tools --with-pic
-    INSTALL_COMMAND make install
+    URL_HASH MD5=29117bf488887f89888f9304c8ebea0b
+    CMAKE_ARGS
+       -DCMAKE_INSTALL_PREFIX=${EXT_INSTALL}
+       -DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
+       -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
+       -DHDF5_BUILD_CPP_LIB=ON
+       -DHDF5_BUILD_TOOLS=ON
+       -DHDF5_BUILD_HL_LIB=ON
+       -DBUILD_SHARED_LIBS=OFF
  )
  SET(cosmotool_DEPS ${cosmotool_DEPS} hdf5)
  SET(hdf5_built hdf5)
  SET(ENV{HDF5_ROOT} ${HDF5_BIN_DIR})
  SET(HDF5_ROOTDIR ${HDF5_BIN_DIR})
  SET(CONFIGURE_LDFLAGS "${CONFIGURE_LDFLAGS} -L${HDF5_BIN_DIR}/lib")
-  SET(CONFIGURE_LIBS "${CONFIGURE_LIBS} -ldl ${RT_LIBRARY}")
-  set(HDF5_C_STATIC_LIBRARY ${HDF5_BIN_DIR}/lib/libhdf5.a)
-  set(HDF5_HL_STATIC_LIBRARY ${HDF5_BIN_DIR}/lib/libhdf5_hl.a)
-  set(HDF5_LIBRARIES ${HDF5_BIN_DIR}/lib/libhdf5.a CACHE STRING "HDF5 lib" FORCE)
-  set(HDF5_HL_LIBRARIES ${HDF5_BIN_DIR}/lib/libhdf5_hl.a CACHE STRING "HDF5 HL lib" FORCE)
-  set(HDF5_CXX_LIBRARIES ${HDF5_BIN_DIR}/lib/libhdf5_cpp.a CACHE STRING "HDF5 C++ lib" FORCE)
+  SET(CONFIGURE_LIBS "${CONFIGURE_LIBS} -ldl")
+  set(HDF5_C_STATIC_LIBRARY ${HDF5_BIN_DIR}/lib/libhdf5-static.a)
+  set(HDF5_HL_STATIC_LIBRARY ${HDF5_BIN_DIR}/lib/libhdf5_hl-static.a)
+  set(HDF5_LIBRARIES ${HDF5_BIN_DIR}/lib/libhdf5-static.a CACHE STRING "HDF5 lib" FORCE)
+  set(HDF5_HL_LIBRARIES ${HDF5_BIN_DIR}/lib/libhdf5_hl-static.a CACHE STRING "HDF5 HL lib" FORCE)
+  set(HDF5_CXX_LIBRARIES ${HDF5_BIN_DIR}/lib/libhdf5_cpp-static.a CACHE STRING "HDF5 C++ lib" FORCE)
  SET(HDF5_INCLUDE_DIRS ${HDF5_BIN_DIR}/include CACHE STRING "HDF5 include path" FORCE)
  mark_as_advanced(HDF5_LIBRARIES HDF5_CXX_LIBRARIES HDF5_INCLUDE_DIRS)

@ -168,52 +160,43 @@ if (INTERNAL_NETCDF)
  SET(NETCDF_CONFIG_COMMAND ${NETCDF_SOURCE_DIR}/configure
         --prefix=${NETCDF_BIN_DIR} --libdir=${NETCDF_BIN_DIR}/lib
         --enable-netcdf-4  --with-pic --disable-shared --disable-dap 
-         --disable-byterange --disable-cdmremote --disable-rpc --enable-cxx-4 
+         --disable-cdmremote --disable-rpc --enable-cxx-4 
         --disable-examples ${EXTRA_NC_FLAGS} CC=${CMAKE_C_COMPILER}
         CXX=${CMAKE_CXX_COMPILER})
  list(INSERT CMAKE_PREFIX_PATH 0 ${EXT_INSTALL})
  string(REPLACE ";" "|" CMAKE_PREFIX_PATH_ALT_SEP "${CMAKE_PREFIX_PATH}")
  ExternalProject_Add(netcdf
    DEPENDS ${hdf5_built}
-    URL_HASH MD5=f48ee01534365006934f0c63d4055ea0
    PREFIX ${BUILD_PREFIX}/netcdf-prefix
-    DOWNLOAD_DIR ${SOURCE_PREFIX}/downloads
    URL ${NETCDF_URL}
-    LIST_SEPARATOR |
-    CMAKE_ARGS
+  LIST_SEPARATOR |
+	CMAKE_ARGS
 		-DCMAKE_PREFIX_PATH=${CMAKE_PREFIX_PATH_ALT_SEP}
 		-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
 		-DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
-                -DNC_EXTRA_DEPS=${RT_DEP}
 		-DBUILD_SHARED_LIBS=OFF
 		-DBUILD_TESTING=OFF
 		-DCMAKE_BUILD_TYPE=Release
 		-DENABLE_NETCDF4=ON
-                -DENABLE_BYTERANGE=FALSE
-		-DCMAKE_POSITION_INDEPENDENT_CODE=ON
-		-DENABLE_DAP=OFF
+    -DCMAKE_POSITION_INDEPENDENT_CODE=ON
+    -DENABLE_DAP=OFF
 		-DCMAKE_INSTALL_PREFIX=${NETCDF_BIN_DIR}
                -DHDF5_C_LIBRARY=${HDF5_C_STATIC_LIBRARY}
                -DHDF5_HL_LIBRARY=${HDF5_HL_STATIC_LIBRARY}
                -DHDF5_INCLUDE_DIR=${HDF5_INCLUDE_DIRS}
-		-DCMAKE_INSTALL_LIBDIR=lib
+    -DCMAKE_INSTALL_LIBDIR=lib
  )
  
  SET(NETCDFCXX_SOURCE_DIR ${BUILD_PREFIX}/netcdf-c++-prefix/src/netcdf-c++)
  ExternalProject_Add(netcdf-c++
    DEPENDS ${hdf5_built} netcdf
    PREFIX ${BUILD_PREFIX}/netcdf-c++-prefix
-    DOWNLOAD_DIR ${SOURCE_PREFIX}/downloads
    URL ${NETCDFCXX_URL}
 	CMAKE_ARGS
-		-DCMAKE_PREFIX_PATH=${CMAKE_PREFIX_PATH_ALT_SEP}
 		-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
 		-DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER}
 		-DBUILD_SHARED_LIBS=OFF
-                -DCMAKE_POSITION_INDEPENDENT_CODE=ON
-                -DHDF5_C_LIBRARY=${HDF5_C_STATIC_LIBRARY}
-                -DHDF5_HL_LIBRARY=${HDF5_HL_STATIC_LIBRARY}
-                -DHDF5_INCLUDE_DIR=${HDF5_INCLUDE_DIRS}
+    -DCMAKE_POSITION_INDEPENDENT_CODE=ON
 		-DBUILD_TESTING=OFF
 		-DCMAKE_BUILD_TYPE=Release
 		-DCMAKE_INSTALL_PREFIX=${NETCDF_BIN_DIR}
@ -251,8 +234,7 @@ if (INTERNAL_BOOST)
  ExternalProject_Add(boost
    URL ${BOOST_URL}
    PREFIX ${BUILD_PREFIX}/boost-prefix
-    DOWNLOAD_DIR ${SOURCE_PREFIX}/downloads
-    URL_HASH MD5=f7050f554a65f6a42ece221eaeec1660
+    URL_HASH MD5=da07ca30dd1c0d1fdedbd487efee01bd
    CONFIGURE_COMMAND 
           ${BOOST_SOURCE_DIR}/bootstrap.sh --prefix=${CMAKE_BINARY_DIR}/ext_build/boost
    BUILD_IN_SOURCE 1
@ -295,7 +277,6 @@ IF(INTERNAL_GSL)
  ExternalProject_Add(gsl
    URL ${GSL_URL}
    PREFIX ${BUILD_PREFIX}/gsl-prefix
-    DOWNLOAD_DIR ${SOURCE_PREFIX}/downloads
    CONFIGURE_COMMAND ${GSL_SOURCE_DIR}/configure
           --prefix=${EXT_INSTALL} --disable-shared
           --with-pic
@ -348,7 +329,6 @@ IF(INTERNAL_FFTW)
  ExternalProject_Add(fftw
     URL ${FFTW_URL}
     PREFIX ${BUILD_PREFIX}/fftw-prefix
-    DOWNLOAD_DIR ${SOURCE_PREFIX}/downloads
     CONFIGURE_COMMAND 
           ${FFTW_SOURCE}/configure
                 --prefix=${EXT_INSTALL}
@ -378,7 +358,6 @@ IF (INTERNAL_EIGEN)
  ExternalProject_Add(eigen
     URL ${EIGEN_URL}
     URL_HASH MD5=b9e98a200d2455f06db9c661c5610496
-    DOWNLOAD_DIR ${SOURCE_PREFIX}/downloads
     PREFIX ${BUILD_PREFIX}/eigen-prefix
     CMAKE_ARGS -DCMAKE_INSTALL_PREFIX=${EXT_INSTALL}
       -DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
@ -411,7 +390,6 @@ SET(cosmotool_DEPS ${cosmotool_DEPS} omptl)
 SET(OMPTL_BUILD_DIR ${BUILD_PREFIX}/omptl-prefix/src/omptl)
 ExternalProject_Add(omptl
  PREFIX ${BUILD_PREFIX}/omptl-prefix
-  DOWNLOAD_DIR ${SOURCE_PREFIX}/downloads
  URL ${CMAKE_SOURCE_DIR}/external/omptl-20120422.tar.bz2
  CONFIGURE_COMMAND echo "No configure"
  BUILD_COMMAND echo "No build"
--- a/external/libsharp-8d51946.tar.gz
+++ b/external/libsharp-8d51946.tar.gz
--- a/external/patch-omptl
+++ b/external/patch-omptl
@ -1,6 +1,6 @@
 diff -ur omptl.old/omptl_algorithm omptl/omptl_algorithm
--- omptl.old/omptl_algorithm	2022-06-19 08:21:39.815498672 +0200
-+++ omptl/omptl_algorithm	2022-06-19 08:21:52.953544672 +0200
+--- omptl.old/omptl_algorithm	2012-04-22 16:29:41.000000000 +0200
+++ omptl/omptl_algorithm	2021-06-20 15:40:29.000000000 +0200
@@ -20,7 +20,7 @@
 #define OMPTL_ALGORITHM 1
 
@ -23,13 +23,11 @@ diff -ur omptl.old/omptl_algorithm omptl/omptl_algorithm
 
 #endif /* OMPTL_ALGORITHM */
 diff -ur omptl.old/omptl_algorithm_par.h omptl/omptl_algorithm_par.h
--- omptl.old/omptl_algorithm_par.h	2022-06-19 08:21:39.816498675 +0200
-+++ omptl/omptl_algorithm_par.h	2022-06-19 08:23:50.705956964 +0200
-@@ -20,9 +20,10 @@
- #include <utility>
+--- omptl.old/omptl_algorithm_par.h	2012-04-22 16:29:41.000000000 +0200
+++ omptl/omptl_algorithm_par.h	2021-06-20 15:40:50.000000000 +0200
+@@ -21,8 +21,8 @@
 #include <cmath>
 #include <cstdlib>
-+#include <random>
 
 -#include <omptl/omptl_tools.h>
 -#include <omptl/omptl_numeric>
@ -38,78 +36,7 @@ diff -ur omptl.old/omptl_algorithm_par.h omptl/omptl_algorithm_par.h
 
 #include <iterator>
 
-@@ -510,7 +511,7 @@
- 
- 		typename std::vector<Iterator>::iterator result =
- 		    std::find_if(results.begin(),results.end(),
-			std::bind2nd(std::not_equal_to<Iterator>(), last) );
-+			std::bind(std::not_equal_to<Iterator>(), std::placeholders::_1, last) );
- 
- 		if ( result != results.end() )
- 			return *result;
-@@ -569,7 +570,7 @@
- 
- 		const typename std::vector<Iterator>::iterator result
- 			 = std::find_if(results.begin(), results.end(),
-			std::bind2nd(std::not_equal_to<Iterator>(), last) );
-+			std::bind(std::not_equal_to<Iterator>(), std::placeholders::_1, last) );
- 
- 		if ( result != results.end() )
- 			return *result;
-@@ -654,7 +655,7 @@
- 
- 		const typename std::vector<Iterator>::iterator
- 		result = std::find_if(results.begin(), results.end(),
-					std::bind2nd(std::not_equal_to<Iterator>(), last1));
-+					std::bind(std::not_equal_to<Iterator>(), std::placeholders::_1, last1));
- 
- 		if ( result != results.end() )
- 			return *result;
-@@ -953,7 +954,7 @@
- 		results[t] = std::lower_bound(partitions[t].first, partitions[t].second, value, comp);
- 
- 	const typename std::vector<ForwardIterator>::iterator result =
-		std::find_if(results.begin(), results.end(), std::bind2nd(std::not_equal_to<ForwardIterator>(), last) );
-+		std::find_if(results.begin(), results.end(), std::bind(std::not_equal_to<ForwardIterator>(), std::placeholders::_1, last) );
- 
- 	if (result != results.end())
- 		return *result;
-@@ -1179,7 +1180,7 @@
- 
- namespace detail
- {
-	
-+
- template<typename Iterator, class StrictWeakOrdering>
- Iterator _pivot_range(Iterator first, Iterator last,
- 	const typename std::iterator_traits<Iterator>::value_type pivot,
-@@ -1309,14 +1310,14 @@
- void random_shuffle(RandomAccessIterator first, RandomAccessIterator last,
-                     const unsigned P)
- {
-	std::random_shuffle(first, last);
-+	std::shuffle(first, last, std::mt19937(std::random_device()()));
- }
- 
- template <class RandomAccessIterator, class RandomNumberGenerator>
- void random_shuffle(RandomAccessIterator first, RandomAccessIterator last,
-                     RandomNumberGenerator& rgen, const unsigned P)
- {
-	std::random_shuffle(first, last, rgen);
-+	std::shuffle(first, last, std::mt19937(std::random_device()()));
- }
- 
- // Not (yet) parallelized, not straightforward due to possible dependencies
-@@ -1472,7 +1473,7 @@
-                             const T& new_value, const unsigned P)
- {
- 	return ::omptl::replace_copy_if(first, last, result,
-			std::bind2nd(std::equal_to<T>(), old_value), new_value, P);
-+			std::bind(std::equal_to<T>(), std::placeholders::_1, old_value), new_value, P);
- }
- 
- template <class ForwardIterator, class Predicate, class T>
-@@ -1700,7 +1701,7 @@
+@@ -1700,7 +1700,7 @@
 
 	std::vector<char> pivot_used(pivots.size(), false); // can't be bool due to parallel write
 
@ -118,56 +45,9 @@ diff -ur omptl.old/omptl_algorithm_par.h omptl/omptl_algorithm_par.h
 	assert(1u << max_depth <= P);
 	for (unsigned i = 0; i < max_depth; ++i)
 	{
-@@ -1781,7 +1782,7 @@
- std::cout  << std::endl;
- 
- std::cout << borders.size() << " " << partitions.size() << " " << P << std::endl;
-*/	
-+*/
- 	// Round one: sort final partitions, split remaining
- 	#pragma omp parallel for
- 	for (int i = 0; i < int(partitions.size()); ++i)
-@@ -1814,7 +1815,7 @@
- 
- 			const RandomAccessIterator begin  = partitions[i].first;
- 			const RandomAccessIterator end    = partitions[i].second;
-			
-+
- 			const RandomAccessIterator middle =
- 				detail::_pivot_range(begin, end, pivots[pivot_index], comp);
- 			partitions[i  ] = std::make_pair(begin, middle);
-diff -ur omptl.old/omptl_algorithm_ser.h omptl/omptl_algorithm_ser.h
--- omptl.old/omptl_algorithm_ser.h	2022-06-19 08:21:39.815498672 +0200
-+++ omptl/omptl_algorithm_ser.h	2022-06-19 08:21:52.960544697 +0200
-@@ -14,7 +14,7 @@
- // You should have received a copy of the GNU Lesser General Public
- // License along with this library; if not, write to the Free Software
- // Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
-
-+#include <random>
- 
- namespace omptl
- {
-@@ -463,14 +463,14 @@
- void random_shuffle(RandomAccessIterator first, RandomAccessIterator last,
-       const unsigned P)
- {
-	return ::std::random_shuffle(first, last);
-+	return ::std::shuffle(first, last, std::mt19937(std::random_device()()));
- }
- 
- template <class RandomAccessIterator, class RandomNumberGenerator>
- void random_shuffle(RandomAccessIterator first, RandomAccessIterator last,
-                     RandomNumberGenerator &rgen, const unsigned P)
- {
-	return ::std::random_shuffle(first, last, rgen);
-+	return ::std::shuffle(first, last, std::mt19937(std::random_device()()));
- }
- 
- template <class ForwardIterator, class T>
 diff -ur omptl.old/omptl_numeric omptl/omptl_numeric
--- omptl.old/omptl_numeric	2022-06-19 08:21:39.816498675 +0200
-+++ omptl/omptl_numeric	2022-06-19 08:21:52.955544679 +0200
+--- omptl.old/omptl_numeric	2012-04-22 16:29:41.000000000 +0200
+++ omptl/omptl_numeric	2021-06-20 15:40:29.000000000 +0200
@@ -19,7 +19,7 @@
 #define OMPTL_NUMERIC 1
 
@ -193,8 +73,8 @@ diff -ur omptl.old/omptl_numeric omptl/omptl_numeric
 
 #endif /* OMPTL_NUMERIC */
 diff -ur omptl.old/omptl_numeric_extensions.h omptl/omptl_numeric_extensions.h
--- omptl.old/omptl_numeric_extensions.h	2022-06-19 08:21:39.815498672 +0200
-+++ omptl/omptl_numeric_extensions.h	2022-06-19 08:21:52.956544683 +0200
+--- omptl.old/omptl_numeric_extensions.h	2012-04-22 16:29:41.000000000 +0200
+++ omptl/omptl_numeric_extensions.h	2021-06-20 15:40:29.000000000 +0200
@@ -51,9 +51,9 @@
 } // namespace
 
@ -208,8 +88,8 @@ diff -ur omptl.old/omptl_numeric_extensions.h omptl/omptl_numeric_extensions.h
 
 namespace omptl
 diff -ur omptl.old/omptl_numeric_par.h omptl/omptl_numeric_par.h
--- omptl.old/omptl_numeric_par.h	2022-06-19 08:21:39.816498675 +0200
-+++ omptl/omptl_numeric_par.h	2022-06-19 08:21:52.957544686 +0200
+--- omptl.old/omptl_numeric_par.h	2012-04-22 16:29:41.000000000 +0200
+++ omptl/omptl_numeric_par.h	2021-06-20 15:40:29.000000000 +0200
@@ -23,8 +23,8 @@
 #include <functional>
 #include <iterator>
@ -222,8 +102,8 @@ diff -ur omptl.old/omptl_numeric_par.h omptl/omptl_numeric_par.h
 namespace omptl
 {
 diff -ur omptl.old/omptl_tools.h omptl/omptl_tools.h
--- omptl.old/omptl_tools.h	2022-06-19 08:21:39.816498675 +0200
-+++ omptl/omptl_tools.h	2022-06-19 08:21:52.957544686 +0200
+--- omptl.old/omptl_tools.h	2012-04-22 16:29:41.000000000 +0200
+++ omptl/omptl_tools.h	2021-06-20 15:40:42.000000000 +0200
@@ -25,7 +25,7 @@
 #include <climits>
 #include <iterator>
--- a/pyproject.toml
+++ b/pyproject.toml
@ -1,4 +0,0 @@
-[build-system]
-requires = ["setuptools","wheel","cython"]
-build-backend = "setuptools.build_meta"
-
--- a/python/CMakeLists.txt
+++ b/python/CMakeLists.txt
@ -2,7 +2,7 @@ set(CMAKE_SHARED_MODULE_PREFIX)


 set(PYTHON_INCLUDES ${NUMPY_INCLUDE_DIRS} ${PYTHON_INCLUDE_PATH} ${CMAKE_SOURCE_DIR}/python)
-include_directories(${CMAKE_SOURCE_DIR}/python ${CMAKE_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/src)
+include_directories(${CMAKE_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/src)

 IF(CYTHON)
  add_custom_command(
@ -69,10 +69,13 @@ target_link_libraries(_fast_interp PRIVATE ${CosmoTool_local} )
 SET(ct_TARGETS   _cosmotool _project _cosmo_power _cosmo_cic _fast_interp _cosmomath)

 if (Boost_FOUND)
-  message(STATUS "Building bispectrum support")
+  message(STATUS "Building bispectrum support (path = ${Boost_INCLUDE_DIRS})")
  include_directories(${Boost_INCLUDE_DIRS})
  add_library(_cosmo_bispectrum MODULE _cosmo_bispectrum.cpp)
  target_link_libraries(_cosmo_bispectrum ${MATH_LIBRARY})
+  if(ENABLE_OPENMP)
+    set_target_properties(_cosmo_bispectrum PROPERTIES COMPILE_FLAGS "${OpenMP_CXX_FLAGS}" LINK_FLAGS "${OpenMP_CXX_FLAGS}")
+  endif()
  if (Boost_DEP)
    add_dependencies(_cosmo_bispectrum ${Boost_DEP})
  endif()
--- a/python/_cosmomath.pyx
+++ b/python/_cosmomath.pyx
@ -12,7 +12,7 @@ cdef extern from "numpy/npy_common.h":
   ctypedef npy_intp

 cdef extern from "special_math.hpp" namespace "CosmoTool":
-   T log_modified_bessel_first_kind[T](T v, T z) except + nogil
+   T log_modified_bessel_first_kind[T](T v, T z) nogil except +

 cdef extern from "numpy_adaptors.hpp" namespace "CosmoTool":
   void parallel_ufunc_dd_d[T,IT](char **args, IT* dimensions, IT* steps, void *func)
--- a/python/_cosmotool.pyx
+++ b/python/_cosmotool.pyx
@ -39,7 +39,7 @@ cdef extern from "loadSimu.hpp" namespace "CosmoTool":

 cdef extern from "loadGadget.hpp" namespace "CosmoTool":

-  SimuData *loadGadgetMulti(const char *fname, int id, int flags, int gformat) except + nogil
+  SimuData *loadGadgetMulti(const char *fname, int id, int flags, int gformat) nogil except +
  void cxx_writeGadget "CosmoTool::writeGadget" (const char * s, SimuData *data) except +

 cdef extern from "safe_gadget.hpp":
@ -313,13 +313,6 @@ tries to get an array from the object."""
 references to the object are gone. """
        pass

-
-cdef extern from "numpy/arrayobject.h":
-    # a little bit awkward: the reference to obj will be stolen
-    # using PyObject*  to signal that Cython cannot handle it automatically
-    int PyArray_SetBaseObject(np.ndarray arr, PyObject *obj) except -1 # -1 means there was an error
-
-
 cdef object wrap_array(void *p, np.uint64_t s, int typ):
    cdef np.ndarray ndarray
    cdef ArrayWrapper wrapper
@ -327,8 +320,7 @@ cdef object wrap_array(void *p, np.uint64_t s, int typ):
    wrapper = ArrayWrapper()
    wrapper.set_data(s, typ, p)
    ndarray = np.array(wrapper, copy=False)
-    #ndarray.base = <PyObject*> wrapper
-    PyArray_SetBaseObject(ndarray, <PyObject*> wrapper)
+    ndarray.base = <PyObject*> wrapper
    Py_INCREF(wrapper)

    return ndarray
--- a/python/_project.pyx
+++ b/python/_project.pyx
@ -727,7 +727,7 @@ cdef DTYPE_t C_line_of_sight_projection(DTYPE_t[:,:,:] density,
    cdef int iu0[3]
    cdef int i
    cdef int N = density.shape[0]
-    cdef int half_N = density.shape[0]//2
+    cdef int half_N = density.shape[0]/2
    cdef int completed
    cdef DTYPE_t I0, d, dist2, delta, s, max_distance2
    cdef int jumper[1]
--- a/requirements.txt
+++ b/requirements.txt
@ -1,4 +1,5 @@
-numpy<1.22
+numpy<1.19
 cffi
 numexpr
-cython<3
+pyfftw
+cython
--- a/sample/CMakeLists.txt
+++ b/sample/CMakeLists.txt
@ -1,8 +1,4 @@
 SET(tolink  ${CosmoTool_local} ${CosmoTool_LIBS} ${GSL_LIBRARIES} ${DL_LIBRARY})
-if (RT_LIBRARY)
-  SET(tolink ${tolink} ${RT_LIBRARY})
-endif()
-
 include_directories(${CMAKE_SOURCE_DIR}/src)
 include_directories(${FFTW3_INCLUDE_DIRS} ${EIGEN3_INCLUDE_DIRS} ${GSL_INCLUDE_PATH})
 if(YORICK_SUPPORT)
@ -91,7 +87,7 @@ if (Boost_FOUND)
  ENDIF (YORICK_SUPPORT)
  if (HDF5_FOUND)
    add_executable(simple3DFilter simple3DFilter.cpp)
-    target_link_libraries(simple3DFilter ${tolink})
+    target_link_libraries(simple3DFilter ${tolink}  ${MPI_C_LIBRARIES})

    add_executable(simpleDistanceFilter simpleDistanceFilter.cpp)
    target_link_libraries(simpleDistanceFilter ${tolink})
--- a/sample/simple3DFilter.cpp
+++ b/sample/simple3DFilter.cpp
@ -1,3 +1,11 @@
+#include <H5Cpp.h>
+#include <mpi.h>
+
+#include <boost/bind.hpp>
+#include <boost/format.hpp>
+#include <cassert>
+#include <iostream>
+
 #include "hdf5_array.hpp"
 #include "miniargs.hpp"
 #include "mykdtree.hpp"
@ -5,11 +13,6 @@
 #include "openmp.hpp"
 #include "sphSmooth.hpp"
 #include "yorick.hpp"
-#include <H5Cpp.h>
-#include <boost/bind.hpp>
-#include <boost/format.hpp>
-#include <cassert>
-#include <iostream>

 using namespace std;
 using namespace CosmoTool;
@ -27,37 +30,32 @@ typedef boost::multi_array<float, 2> array_type;
 typedef boost::multi_array<float, 3> array3_type;
 typedef boost::multi_array<float, 4> array4_type;

-ComputePrecision getVelocity(const VCoord &v, int i) { return v.mass * v.v[i]; }
+ComputePrecision getVelocity(const VCoord& v, int i) { return v.mass * v.v[i]; }

-ComputePrecision getMass(const VCoord &v) { return v.mass; }
+ComputePrecision getMass(const VCoord& v) { return v.mass; }

 typedef SPHSmooth<VCoord> MySmooth;
 typedef MySmooth::SPHTree MyTree;
 typedef MyTree::Cell MyCell;

 template <typename FuncT>
-void computeInterpolatedField(MyTree *tree1, double boxsize, int Nres,
+void computeInterpolatedField(MyTree* tree1, double boxsize, int Nres,
                              double cx, double cy, double cz,
-                              array3_type &bins, array3_type &arr, FuncT func,
+                              array3_type& bins, array3_type& arr, FuncT func,
                              double rLimit2) {
-  int rz_max = 0;
-#pragma omp parallel shared(rz_max)
+#pragma omp parallel
  {
    MySmooth smooth1(tree1, N_SPH);

-#pragma omp for collapse(3) schedule(dynamic)
+#pragma omp for schedule(dynamic)
    for (int rz = 0; rz < Nres; rz++) {
+      double pz = (rz)*boxsize / Nres - cz;

+      cout << format("[%d] %d / %d") % smp_get_thread_id() % rz % Nres << endl;
      for (int ry = 0; ry < Nres; ry++) {
+        double py = (ry)*boxsize / Nres - cy;
        for (int rx = 0; rx < Nres; rx++) {
-          if (rz > rz_max) {
-            rz_max = rz;
-            cout << format("[%d] %d / %d") % smp_get_thread_id() % rz % Nres
-                 << endl;
-          }
          double px = (rx)*boxsize / Nres - cx;
-          double py = (ry)*boxsize / Nres - cy;
-          double pz = (rz)*boxsize / Nres - cz;

          MyTree::coords c = {float(px), float(py), float(pz)};

@ -80,11 +78,18 @@ void computeInterpolatedField(MyTree *tree1, double boxsize, int Nres,
  }
 }

-int main(int argc, char **argv) {
-  char *fname1, *outFile;
+int main(int argc, char** argv) {
+  int provided;
+  MPI_Init_thread(&argc, &argv, MPI_THREAD_FUNNELED, &provided);
+  if (provided < MPI_THREAD_FUNNELED) {
+    std::cerr << "Cannot mix MPI and Threads here. Please recompile with "
+                 "OpenMP or MPI switched off."
+              << std::endl;
+    MPI_Abort(MPI_COMM_WORLD, 99);
+  }
+  char *fname1, *fname2;
  double rLimit, boxsize, rLimit2, cx, cy, cz;
  int Nres;
-  int periodic;

  MiniArgDesc args[] = {{"INPUT DATA1", &fname1, MINIARG_STRING},
                        {"RADIUS LIMIT", &rLimit, MINIARG_DOUBLE},
@ -93,17 +98,17 @@ int main(int argc, char **argv) {
                        {"CX", &cx, MINIARG_DOUBLE},
                        {"CY", &cy, MINIARG_DOUBLE},
                        {"CZ", &cz, MINIARG_DOUBLE},
-                        {"OUTPUT FILE", &outFile, MINIARG_STRING},
-                        {"PERIODIC", &periodic, MINIARG_INT},
                        {0, 0, MINIARG_NULL}};

-  if (!parseMiniArgs(argc, argv, args))
-    return 1;
+  if (!parseMiniArgs(argc, argv, args)) return 1;

+  int rank, size;
+  MPI_Comm_rank(MPI_COMM_WORLD, &rank);
+  MPI_Comm_size(MPI_COMM_WORLD, &size);
  H5::H5File in_f(fname1, 0);
-  H5::H5File out_f(outFile, H5F_ACC_TRUNC);
+  H5::H5File out_f(boost::str(boost::format("fields_%d.h5", rank), H5F_ACC_TRUNC);
  array_type v1_data;
-  uint64_t N1_points, N2_points;
+  uint32_t N1_points, N2_points;

  array3_type bins(boost::extents[Nres][Nres][Nres]);

@ -116,17 +121,15 @@ int main(int argc, char **argv) {

  cout << "Got " << N1_points << " in the first file." << endl;

-  MyCell *allCells_1 = new MyCell[N1_points];
+  MyCell* allCells_1 = new MyCell[N1_points];

 #pragma omp parallel for schedule(static)
-  for (uint32_t i = 0; i < Nres * Nres * Nres; i++)
-    bins.data()[i] = 0;
+  for (long i = 0; i < Nres * Nres * Nres; i++) bins.data()[i] = 0;

  cout << "Shuffling data in cells..." << endl;
 #pragma omp parallel for schedule(static)
-  for (uint64_t i = 0; i < N1_points; i++) {
-    for (int j = 0; j < 3; j++)
-      allCells_1[i].coord[j] = v1_data[i][j];
+  for (int i = 0; i < N1_points; i++) {
+    for (int j = 0; j < 3; j++) allCells_1[i].coord[j] = v1_data[i][j];
    for (int k = 0; k < 3; k++)
      allCells_1[i].val.pValue.v[k] = v1_data[i][3 + k];
    allCells_1[i].val.pValue.mass = v1_data[i][6];
@ -140,9 +143,8 @@ int main(int argc, char **argv) {
    if (rx < 0 || rx >= Nres || ry < 0 || ry >= Nres || rz < 0 || rz >= Nres)
      continue;

-    auto &b = bins[rx][ry][rz];
-#pragma omp atomic
-    b++;
+    //#pragma omp atomic update
+    bins[rx][ry][rz]++;
  }
  v1_data.resize(boost::extents[1][1]);

@ -150,7 +152,6 @@ int main(int argc, char **argv) {

  cout << "Building trees..." << endl;
  MyTree tree1(allCells_1, N1_points);
-  tree1.setPeriodic(periodic != 0, boxsize);

  cout << "Creating smoothing filter..." << endl;

@ -158,21 +159,18 @@ int main(int argc, char **argv) {

  cout << "Weighing..." << endl;

-  int rz_max = 0;
-#pragma omp parallel shared(rz_max)
+#pragma omp parallel
  {
    MySmooth smooth1(&tree1, N_SPH);

-#pragma omp for collapse(3) schedule(dynamic, 8)
+#pragma omp for schedule(dynamic)
    for (int rz = 0; rz < Nres; rz++) {
+      double pz = (rz)*boxsize / Nres - cz;
+
+      (cout << rz << " / " << Nres << endl).flush();
      for (int ry = 0; ry < Nres; ry++) {
+        double py = (ry)*boxsize / Nres - cy;
        for (int rx = 0; rx < Nres; rx++) {
-          if (rz > rz_max) {
-            rz_max = rz;
-            (cout << rz << " / " << Nres << endl).flush();
-          }
-          double pz = (rz)*boxsize / Nres - cz;
-          double py = (ry)*boxsize / Nres - cy;
          double px = (rx)*boxsize / Nres - cx;

          MyTree::coords c = {float(px), float(py), float(pz)};
@ -187,11 +185,30 @@ int main(int argc, char **argv) {
            smooth1.fetchNeighbours(c, numInCell);
          else
            smooth1.fetchNeighbours(c);
+#pragma omp critical
          smooth1.addGridSite(c);
        }
      }
+      (cout << " Done " << rz << endl).flush();
    }
  }
+  //
+  // Reduction on the cell.weight in the tree.
+  // MPI_Allreduce to act on contiguous arrays.
+  auto tree = smooth1.getTree();
+  auto nodes = tree->getAllNodes();
+  double *weight_array = new double[N1_points];
+  for (size_t c = 0; c < N1_points; c++) {
+    weight_array[c] = allCells[c].val.weight;
+  }
+  MPI_Allreduce(MPI_IN_PLACE, weight_array, N1_points, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  for (size_t c = 0; c < N1_points; c++) {
+    allCells[c].val.weight = weight_array[c];
+  }
+  delete[] weight_array;
+
+  // cell.weights  -> build a 1d array of the particles weight -> MPI_Allreduce -> resend the new weights to the particles
+  //

  cout << "Interpolating..." << endl;

--- a/setup.py
+++ b/setup.py
@ -2,7 +2,6 @@ import stat
 import os
 import sys
 import shutil
-from sysconfig import get_config_var
 from distutils.command.install_data import install_data
 import pathlib
 from setuptools import find_packages, setup, Extension
@ -168,24 +167,17 @@ class BuildCMakeExt(build_ext):
        # Change your cmake arguments below as necessary
        # Below is just an example set of arguments for building Blender as a Python module

-        c_compiler=os.environ.get('CC', get_config_var("CC"))
-        cxx_compiler=os.environ.get('CXX', get_config_var("CXX"))
-
        compilers=[]
        fill_up_settings=[
-              ("CMAKE_C_COMPILER", "CC", get_config_var("CC")),
-              ("CMAKE_CXX_COMPILER", "CXX", get_config_var("CXX")),
-              ("CMAKE_C_FLAGS", "CFLAGS", None),
-              ("CMAKE_CXX_FLAGS", "CXXFLAGS", None),
-              ("CMAKE_EXE_LINKER_FLAGS_INIT", "LDFLAGS", None),
-              ("CMAKE_SHARED_LINKER_FLAGS_INIT", "LDFLAGS", None),
-              ("CMAKE_MODULE_LINKER_FLAGS_INIT", "LDFLAGS", None)]
-        for cmake_flag, os_flag, default_flag in fill_up_settings:
-          if os_flag in os.environ or default_flag is not None:
-             c = os.environ.get(os_flag, default_flag)
-             if not c is None:
-               compilers.append(f"-D{cmake_flag}={c}")
-        print(compilers)
+              ("CMAKE_C_COMPILER", "CC"),
+              ("CMAKE_CXX_COMPILER", "CXX"),
+              ("CMAKE_C_FLAGS", "CFLAGS"),
+              ("CMAKE_EXE_LINKER_FLAGS_INIT", "LDFLAGS"),
+              ("CMAKE_SHARED_LINKER_FLAGS_INIT", "LDFLAGS"),
+              ("CMAKE_MODULE_LINKER_FLAGS_INIT", "LDFLAGS")]
+        for cmake_flag, os_flag in fill_up_settings:
+          if os_flag in os.environ:
+             compilers.append(f"-D{cmake_flag}={os.environ[os_flag]}")

        self.spawn(['cmake', '-H'+SOURCE_DIR, '-B'+self.build_temp,
                    '-DENABLE_OPENMP=ON','-DINTERNAL_BOOST=ON','-DINTERNAL_EIGEN=ON',
@ -237,10 +229,10 @@ class BuildCMakeExt(build_ext):
 CosmoTool_extension = CMakeExtension(name="cosmotool")

 setup(name='cosmotool',
-      version='1.3.6',
+      version='1.2.3',
      packages=["cosmotool"],
      package_dir={'cosmotool': 'python/cosmotool'},
-      install_requires=['cython','numpy','cffi','numexpr','h5py'],
+      install_requires=['numpy','cffi','numexpr','pyfftw','h5py'],
      setup_requires=['cython','cffi','numpy','numexpr'],
      ext_modules=[CosmoTool_extension],
      description='A small cosmotool box of useful functions',
--- a/src/cosmopower.cpp
+++ b/src/cosmopower.cpp
@ -212,30 +212,6 @@ double CosmoPower::powerHuWiggles(double k)
  return normPower * pow(k,n) * T_k * T_k;
 }

-double CosmoPower::BAO_Tk(double k){
-
-  double no_wiggle_tk = noWiggleTk(k);
-
-  double A = 0;
-  double r_s = 10;
-  double k_D = 2 * M_PI / 100;
-
-  //sqrt as we want to make the parameterization part of the transfer function
-  double param = sqrt(1 + A * sin(k * r_s) * exp(- k / k_D)); 
-  return no_wiggle_tk * param;
-
-}
-
-double CosmoPower::sample_BAO(double k)
-{
-  // BAO wiggle parameterization for reconstruction
-  // Babic et al. 2022, https://arxiv.org/abs/2203.06177
-
-  double T_k = BAO_Tk(k); 
-
-  return normPower * pow(k,n) * T_k * T_k;
-  }
-
 double CosmoPower::primordialPowerSpectrum(double k)
 {
  //Primordial power spectrum, needed for PNG
@ -273,19 +249,6 @@ double CosmoPower::matterTransferFunctionHu(double k)
  return T_k;
 }

-double CosmoPower::noWiggleTk(double k)
-{
-  double s = 44.5 * log(9.83 / OmegaEff) / (sqrt(1 + 10 * pow(OMEGA_B * h * h, 0.75)));
-  double alpha_Gamma = 1 - 0.328 * log(431 * OmegaEff) * OMEGA_B / OMEGA_0 + 0.38 * log(22.3 * OmegaEff) * pow(OMEGA_B / OMEGA_0, 2);
-  double GammaEff = OMEGA_0 * h * (alpha_Gamma + (1 - alpha_Gamma)/(1 + pow(0.43 * k * s, 4)));
-  double q = k/(h*GammaEff) * pow(Theta_27, 2);
-  double L_0 = log(2 * M_E + 1.8 * q);
-  double C_0 = 14.2 + 731 / (1 + 62.5 * q);
-  double T0 = L_0 / (L_0 + C_0 * q * q);
-
-  return T0;
-}
-
 double CosmoPower::powerHuBaryons(double k)
 {
  double s = 44.5 * log(9.83 / OmegaEff) / (sqrt(1 + 10 * pow(OMEGA_B * h * h, 0.75)));
@ -481,18 +444,12 @@ void CosmoPower::setFunction(CosmoFunction f)
    case POWER_SUGIYAMA:
      eval = &CosmoPower::powerSugiyama;
      break;
-    case SAMPLE_WIGGLES:
-      eval = &CosmoPower::sample_BAO;
-      break;
    case POWER_BDM:
      eval = &CosmoPower::powerBDM;
      break;
    case POWER_TEST:
      eval = &CosmoPower::powerTest;
      break;
-    case BAO_TK:
-      eval = &CosmoPower::BAO_Tk;
-      break;
    default:
      abort();
    }
--- a/src/cosmopower.hpp
+++ b/src/cosmopower.hpp
@ -89,9 +89,7 @@ namespace CosmoTool {
      POWER_SUGIYAMA,
      POWER_BDM,
      POWER_TEST,
-      HU_WIGGLES_ORIGINAL,
-      SAMPLE_WIGGLES,
-      BAO_TK
+      HU_WIGGLES_ORIGINAL
    };

    CosmoPower();
@ -124,9 +122,6 @@ namespace CosmoTool {
    double powerBDM(double k);
    double powerTest(double k);
    double powerHuWigglesOriginal(double k);
-    double sample_BAO(double k);
-    double noWiggleTk(double k);
-    double BAO_Tk(double k);
  };

 };
--- a/src/h5_readFlash.cpp
+++ b/src/h5_readFlash.cpp
@ -123,7 +123,10 @@ void h5_read_runtime_parameters
  DataSpace memspace(rank, &dimens_1d);
  //memspace = H5Screate_simple(rank, &dimens_1d, NULL);

-  dataset.read(int_rt_parms, int_rt_type, memspace, dataspace);
+  dataset.read(int_rt_parms, int_rt_type, memspace, dataspace,
+		    H5P_DEFAULT);
+  //status = H5Dread(dataset, int_rt_type, memspace, dataspace,
+	//	    H5P_DEFAULT, int_rt_parms);


  for (i = 0; i < nint_runtime_parameters; i++) {
@ -164,7 +167,10 @@ void h5_read_runtime_parameters
   memspace = DataSpace(rank, &dimens_1d);
  //memspace = H5Screate_simple(rank, &dimens_1d, NULL);

-  dataset.read(real_rt_parms, real_rt_type, memspace, dataspace);
+  dataset.read(real_rt_parms, real_rt_type, memspace, dataspace,
+		    H5P_DEFAULT);
+  //status = H5Dread(dataset, real_rt_type, memspace, dataspace,
+//		    H5P_DEFAULT, real_rt_parms);


  for (i = 0; i < nreal_runtime_parameters; i++) {
@ -283,7 +289,9 @@ void h5_read_flash3_particles (H5File* file,
  DataSpace memspace(rank, dimens_2d);
  //memspace = H5Screate_simple(rank, dimens_2d, NULL);

-  dataset.read(part_names, string_type);
+  dataset.read(part_names, string_type, H5S_ALL, H5S_ALL, H5P_DEFAULT);
+  //status = H5Dread(dataset, string_type, H5S_ALL, H5S_ALL,
+  //              H5P_DEFAULT, part_names);

 
  string_type.close();
@ -366,7 +374,8 @@ void h5_read_flash3_particles (H5File* file,
    //memspace   = H5Screate_simple(rank, dimens_2d, maxdimens_2d);

    /* read data from the dataset */
-   dataset.read(partBuffer, datatype, memspace, dataspace);
+   dataset.read(partBuffer, datatype, memspace, dataspace, H5P_DEFAULT);
+   //status = H5Dread(dataset, datatype, memspace, dataspace, H5P_DEFAULT, partBuffer);

    /* convert buffer into particle struct */

@ -476,7 +485,8 @@ void h5_read_flash3_header_info(H5File* file,
          PredType::NATIVE_DOUBLE);

  // read the data into 'real_list' 
-  dataset.read( real_list, real_list_type, memspace, dataspace);
+  dataset.read( real_list, real_list_type, memspace, dataspace,
+                H5P_DEFAULT);


  if (status < 0) {
--- a/src/hdf5_array.hpp
+++ b/src/hdf5_array.hpp
@ -426,7 +426,7 @@ namespace CosmoTool {
 #define CTOOL_HDF5_INSERT_ELEMENT(r, STRUCT, element) \
     { \
       ::CosmoTool::get_hdf5_data_type<BOOST_PP_TUPLE_ELEM(2, 0, element)> t; \
-       long position = offsetof(STRUCT, BOOST_PP_TUPLE_ELEM(2, 1, element)); \
+       long position = HOFFSET(STRUCT, BOOST_PP_TUPLE_ELEM(2, 1, element)); \
       const char *field_name = BOOST_PP_STRINGIZE(BOOST_PP_TUPLE_ELEM(2, 1, element)); \
       type.insertMember(field_name,  position, t.type()); \
     } 
--- a/src/interpolate.cpp
+++ b/src/interpolate.cpp
@ -157,7 +157,6 @@ const Interpolate& Interpolate::operator=(const Interpolate& a)
  gsl_spline_init(spline, a.spline->x, a.spline->y, a.spline->size);
  logx = a.logx;
  logy = a.logy;
-  return *this;
 }

 double Interpolate::getMaxX() const
--- a/src/mykdtree.hpp
+++ b/src/mykdtree.hpp
@ -1,5 +1,5 @@
 /*+
-This is CosmoTool (./src/mykdtree.hpp) -- Copyright (C) Guilhem Lavaux (2007-2022)
+This is CosmoTool (./src/mykdtree.hpp) -- Copyright (C) Guilhem Lavaux (2007-2014)

 guilhem.lavaux@gmail.com

@ -99,8 +99,8 @@ namespace CosmoTool {
    typename KDDef<N,CType>::CoordType r, r2;
    KDCell<N, ValType,CType> **cells;
    typename KDDef<N,CType>::CoordType *distances;
-    uint64_t currentRank;
-    uint64_t numCells;
+    uint32_t currentRank;
+    uint32_t numCells;
  };
  

@ -114,7 +114,7 @@ namespace CosmoTool {

    RecursionMultipleInfo(const typename KDDef<N,CType>::KDCoordinates& rx,
 			  KDCell<N,ValType,CType> **cells,
-			  uint64_t numCells)
+			  uint32_t numCells)
      : queue(cells, numCells, INFINITY),traversed(0)
    {
      std::copy(rx, rx+N, x);
@ -153,20 +153,20 @@ namespace CosmoTool {
       std::copy(replicate, replicate+N, this->replicate);
     }

-    uint64_t getIntersection(const coords& x, CoordType r,
+    uint32_t getIntersection(const coords& x, CoordType r, 
 			     Cell **cells,
-			     uint64_t numCells);
-    uint64_t getIntersection(const coords& x, CoordType r,
+			     uint32_t numCells);
+    uint32_t getIntersection(const coords& x, CoordType r, 
 			     Cell **cells,
 			     CoordType *distances,
-			     uint64_t numCells);
-    uint64_t countCells(const coords& x, CoordType r);
+			     uint32_t numCells);
+    uint32_t countCells(const coords& x, CoordType r);

    Cell *getNearestNeighbour(const coords& x);

-    void getNearestNeighbours(const coords& x, uint64_t NumCells,
+    void getNearestNeighbours(const coords& x, uint32_t NumCells,
 			      Cell **cells);
-    void getNearestNeighbours(const coords& x, uint64_t NumCells,
+    void getNearestNeighbours(const coords& x, uint32_t NumCells,
 			      Cell **cells,
 			      CoordType *distances);

--- a/src/mykdtree.tcc
+++ b/src/mykdtree.tcc
@ -80,9 +80,9 @@ namespace CosmoTool {
  }

  template<int N, typename ValType, typename CType, typename CellSplitter>
-  uint64_t KDTree<N,ValType,CType,CellSplitter>::getIntersection(const coords& x, CoordType r,
+  uint32_t KDTree<N,ValType,CType,CellSplitter>::getIntersection(const coords& x, CoordType r, 
 						    KDTree<N,ValType,CType,CellSplitter>::Cell **cells,
-						    uint64_t numCells)
+						    uint32_t numCells)
  {
    RecursionInfoCells<N,ValType,CType> info;

@ -112,10 +112,10 @@ namespace CosmoTool {
  }

  template<int N, typename ValType, typename CType, typename CellSplitter>
-  uint64_t KDTree<N,ValType,CType,CellSplitter>::getIntersection(const coords& x, CoordType r,
+  uint32_t KDTree<N,ValType,CType,CellSplitter>::getIntersection(const coords& x, CoordType r, 
 						    Cell **cells,
 						    CoordType *distances,
-						    uint64_t numCells)
+						    uint32_t numCells)
  {
    RecursionInfoCells<N,ValType> info;

@ -144,7 +144,7 @@ namespace CosmoTool {
  }

  template<int N, typename ValType, typename CType, typename CellSplitter>
-  uint64_t KDTree<N,ValType,CType,CellSplitter>::countCells(const coords& x, CoordType r)
+  uint32_t KDTree<N,ValType,CType,CellSplitter>::countCells(const coords& x, CoordType r)
  {
    RecursionInfoCells<N,ValType> info;

@ -501,7 +501,7 @@ namespace CosmoTool {
  }

 template<int N, typename ValType, typename CType, typename CellSplitter>
- void KDTree<N,ValType,CType,CellSplitter>::getNearestNeighbours(const coords& x, uint64_t N2,
+ void KDTree<N,ValType,CType,CellSplitter>::getNearestNeighbours(const coords& x, uint32_t N2,
 						    Cell **cells)
 {
   RecursionMultipleInfo<N,ValType> info(x, cells, N2);
@ -527,7 +527,7 @@ namespace CosmoTool {
 }

 template<int N, typename ValType, typename CType, typename CellSplitter>
- void KDTree<N,ValType,CType,CellSplitter>::getNearestNeighbours(const coords& x, uint64_t N2,
+ void KDTree<N,ValType,CType,CellSplitter>::getNearestNeighbours(const coords& x, uint32_t N2,
 						    Cell **cells,
 						    CoordType *distances)
 {
@ -643,8 +643,8 @@ namespace CosmoTool {
           throw InvalidOnDiskKDTree();
        }
        if (node_on_disk.cell_id > numNodes || node_on_disk.cell_id < 0 ||
-	    (node_on_disk.children_node[0] >= 0 && node_on_disk.children_node[0] > lastNode) || node_on_disk.children_node[0] < -1 ||
-	    (node_on_disk.children_node[1] >= 0 && node_on_disk.children_node[1] > lastNode) || node_on_disk.children_node[1] < -1)
+	    node_on_disk.children_node[0] > lastNode || node_on_disk.children_node[0] < -1 ||
+	    node_on_disk.children_node[1] > lastNode || node_on_disk.children_node[1] < -1)
 	  {
 	    delete[] nodes;
            std::cerr << "Invalid cell id or children node id invalid" << std::endl;
--- a/src/powerSpectrum.cpp
+++ b/src/powerSpectrum.cpp
@ -58,7 +58,6 @@ using namespace std;
 #define POWER_TEST 8

 #define POWER_SPECTRUM POWER_EFSTATHIOU
-#define SAMPLE_WIGGLES 9

 namespace Cosmology {

@ -207,31 +206,6 @@ double powG(double y)
 */
 double powerSpectrum(double k, double normPower)
 {
-#if POWER_SPECTRUM == SAMPLE_WIGGLES
-  // BAO wiggle parameterization for reconstruction
-  // Babic et al. 2022, https://arxiv.org/abs/2203.06177
-
-  // No-wiggle transfer function
-
-  double s = 44.5 * log(9.83 / OmegaEff) / (sqrt(1 + 10 * pow(OMEGA_B * h * h, 0.75)));
-  double alpha_Gamma = 1 - 0.328 * log(431 * OmegaEff) * OMEGA_B / OMEGA_0 + 0.38 * log(22.3 * OmegaEff) * pow(OMEGA_B / OMEGA_0, 2);
-  double GammaEff = OMEGA_0 * h * (alpha_Gamma + (1 - alpha_Gamma)/(1 + pow(0.43 * k * s, 4)));
-  double q = k/(h*GammaEff) * pow(Theta_27, 2);
-  double L_0 = log(2 * M_E + 1.8 * q);
-  double C_0 = 14.2 + 731 / (1 + 62.5 * q);
-  double T0 = L_0 / (L_0 + C_0 * q * q);
-
-  // Wiggle parameterization
-
-  double A = 0;
-  double r_s = 10;
-  double k_D = 2 * M_PI / 100;
-
-  double param = 1 + A * sin(k * r_s) * exp(- k / k_D);
-
-  return normPower * pow(k, n) * T0 * T0 * param;
-#endif
-
 #if POWER_SPECTRUM == POWER_EFSTATHIOU
  double a = 6.4/Gamma0;
  double b = 3/Gamma0;
--- a/src/sphSmooth.hpp
+++ b/src/sphSmooth.hpp
@ -1,5 +1,5 @@
 /*+
-This is CosmoTool (./src/sphSmooth.hpp) -- Copyright (C) Guilhem Lavaux (2007-2022)
+This is CosmoTool (./src/sphSmooth.hpp) -- Copyright (C) Guilhem Lavaux (2007-2014)

 guilhem.lavaux@gmail.com

@ -39,26 +39,30 @@ knowledge of the CeCILL license and that you accept its terms.
 #include "config.hpp"
 #include "mykdtree.hpp"

-namespace CosmoTool {
+namespace CosmoTool
+{
  template <typename ValType, int Ndims = NUMDIMS>
-  class SPHSmooth {
+  class SPHSmooth
+  {
  public:
-    typedef struct {
+    typedef struct
+    {
      ComputePrecision weight;
      ValType pValue;
    } FullType;

-    typedef KDTree<Ndims, FullType> SPHTree;
-    typedef KDTreeNode<Ndims, FullType> SPHNode;
-    typedef KDCell<Ndims, FullType> SPHCell;
-    typedef typename KDTree<Ndims, FullType>::CoordType CoordType;
+    typedef KDTree<Ndims,FullType> SPHTree;
+    typedef KDTreeNode<Ndims,FullType> SPHNode;
+    typedef KDCell<Ndims,FullType>  SPHCell;
+    typedef typename KDTree<Ndims,FullType>::CoordType CoordType;

-    typedef ComputePrecision (*computeParticleValue)(const ValType &t);
-    typedef void (*runParticleValue)(ValType &t);
+    typedef ComputePrecision (*computeParticleValue)(const ValType& t);
+    typedef void (*runParticleValue)(ValType& t);

  public:
    typedef SPHCell *P_SPHCell;
-    struct SPHState {
+    struct SPHState
+    {
      boost::shared_ptr<P_SPHCell[]> ngb;
      boost::shared_ptr<CoordType[]> distances;
      typename SPHTree::coords currentCenter;
@ -66,58 +70,45 @@ namespace CosmoTool {
      ComputePrecision smoothRadius;
    };
    
+  
    SPHSmooth(SPHTree *tree, uint32_t Nsph);
    virtual ~SPHSmooth();   

-    void
-    fetchNeighbours(const typename SPHTree::coords &c, SPHState *state = 0);
+    void fetchNeighbours(const typename SPHTree::coords& c, SPHState *state = 0);
    
-    void fetchNeighbours(
-        const typename SPHTree::coords &c, uint32_t newNsph,
-        SPHState *state = 0);
-    void fetchNeighboursOnVolume(
-        const typename SPHTree::coords &c, ComputePrecision radius);
-    const typename SPHTree::coords &getCurrentCenter() const {
+    void fetchNeighbours(const typename SPHTree::coords& c, uint32_t newNsph);
+    void fetchNeighboursOnVolume(const typename SPHTree::coords& c, ComputePrecision radius);
+    const typename SPHTree::coords& getCurrentCenter() const
+    {
      return internal.currentCenter;
    }

-    /** This is the pure SPH smoothing function. It does not reweight by the
-     * value computed at each grid site.
-     */
-    template <typename FuncT>
-    ComputePrecision computeSmoothedValue(
-        const typename SPHTree::coords &c, FuncT fun, SPHState *state = 0);
+    template<typename FuncT>
+    ComputePrecision computeSmoothedValue(const typename SPHTree::coords& c,
+					  FuncT fun, SPHState *state = 0);    
 	
-    /** This is the weighted SPH smoothing function. It does reweight by the
-     * value computed at each grid site. This ensures the total sum of the interpolated
-     * quantity is preserved by interpolating to the target mesh.
-     */
-    template <typename FuncT>
-    ComputePrecision computeInterpolatedValue(
-        const typename SPHTree::coords &c, FuncT fun, SPHState *state = 0);
+	template<typename FuncT>
+    ComputePrecision computeInterpolatedValue(const typename SPHTree::coords& c,
+					      FuncT fun, SPHState *state = 0);
 					          
-    /** This is the adjoint gradient of computeInterpolatedValue w.r.t. to the value
-     * array. FuncT is expected to have the following prototype:
-     *    void((CellValue defined by the user), ComputePrecision weighted_ag_value)
-     */
-    template <typename FuncT>
-    void computeAdjointGradientSmoothedValue(
-        const typename SPHTree::coords &c, ComputePrecision ag_value, FuncT fun,
-        SPHState *state = 0);
+    ComputePrecision getMaxDistance(const typename SPHTree::coords& c, 
+				    SPHNode *node) const;

-    ComputePrecision
-    getMaxDistance(const typename SPHTree::coords &c, SPHNode *node) const;
-
-    ComputePrecision getSmoothingLen() const { return internal.smoothRadius; }
+    ComputePrecision getSmoothingLen() const
+    {
+      return internal.smoothRadius;
+    }

    // TO USE WITH EXTREME CARE !
-    void setSmoothingLen(ComputePrecision len) { internal.smoothRadius = len; }
+    void setSmoothingLen(ComputePrecision len)
+    {
+      internal.smoothRadius = len;
+    }
    // END

-    template <typename FuncT>
+    template<typename FuncT>
    void runForEachNeighbour(FuncT fun, SPHState *state = 0);
-    void addGridSite(const typename SPHTree::coords &c, SPHState *state);
-    void addGridSite(const typename SPHTree::coords &c);
+    void addGridSite(const typename SPHTree::coords& c);

    bool hasNeighbours() const;

@ -142,21 +133,21 @@ namespace CosmoTool {
    uint32_t maxNgb;
    SPHTree *tree;
    
-    template <typename FuncT>
-    ComputePrecision computeWValue(
-        const typename SPHTree::coords &c, SPHCell &cell, CoordType d,
-        FuncT fun, SPHState *state);
+    template<typename FuncT>
+    ComputePrecision computeWValue(const typename SPHTree::coords & c,
+				   SPHCell& cell,
+				   CoordType d,
+				   FuncT fun, SPHState *state);
    
-    template <typename FuncT>
-    void runUnrollNode(SPHNode *node, FuncT fun);
+    template<typename FuncT>
+    void runUnrollNode(SPHNode *node,
+		       FuncT fun);
  };

-  template <class ValType1, class ValType2, int Ndims>
-  bool operator<(
-      const SPHSmooth<ValType1, Ndims> &s1,
-      const SPHSmooth<ValType2, Ndims> &s2);
+  template<class ValType1, class ValType2, int Ndims>
+  bool operator<(const SPHSmooth<ValType1, Ndims>& s1, const SPHSmooth<ValType2, Ndims>& s2);

-}; // namespace CosmoTool
+};

 #include "sphSmooth.tcc"

--- a/src/sphSmooth.tcc
+++ b/src/sphSmooth.tcc
@ -3,248 +3,225 @@

 namespace CosmoTool {

-  template <typename ValType, int Ndims>
-  SPHSmooth<ValType, Ndims>::SPHSmooth(SPHTree *tree, uint32_t Nsph) {
-    this->Nsph = Nsph;
-    this->tree = tree;
-    internal.currentNgb = 0;
+template<typename ValType, int Ndims>
+SPHSmooth<ValType,Ndims>::SPHSmooth(SPHTree *tree, uint32_t Nsph)
+{
+  this->Nsph = Nsph;
+  this->tree = tree;
+  internal.currentNgb = 0;

-    this->maxNgb = Nsph;
-    internal.ngb = boost::shared_ptr<SPHCell *[]>(new SPHCell *[maxNgb]);
-    internal.distances = boost::shared_ptr<CoordType[]>(new CoordType[maxNgb]);
-  }
+  this->maxNgb = Nsph;
+  internal.ngb = boost::shared_ptr<SPHCell *[]>(new SPHCell *[maxNgb]);  
+  internal.distances = boost::shared_ptr<CoordType[]>(new CoordType[maxNgb]);
+}

-  template <typename ValType, int Ndims>
-  SPHSmooth<ValType, Ndims>::~SPHSmooth() {}
+template<typename ValType, int Ndims>
+SPHSmooth<ValType,Ndims>::~SPHSmooth()
+{
+}

-  template <typename ValType, int Ndims>
-  template <typename FuncT>
-  ComputePrecision SPHSmooth<ValType, Ndims>::computeWValue(
-      const typename SPHTree::coords &c, SPHCell &cell, CoordType d, FuncT fun,
-      SPHState *state) {
-    CoordType weight;
+template<typename ValType, int Ndims>
+template<typename FuncT>
+ComputePrecision SPHSmooth<ValType,Ndims>::computeWValue(const typename SPHTree::coords& c,
+							 SPHCell& cell,
+							 CoordType d,
+							 FuncT fun, SPHState *state)
+{
+  CoordType weight;

-    d /= state->smoothRadius;
-    weight = getKernel(d);
+  d /= state->smoothRadius;
+  weight = getKernel(d);

-    if (cell.val.weight != 0)
-      return weight * fun(cell.val.pValue) / cell.val.weight;
-    else
-      return 0;
-  }
+  if (cell.val.weight != 0)
+    return weight * fun(cell.val.pValue) / cell.val.weight;
+  else
+    return 0;
+}

-  template <typename ValType, int Ndims>
-  void SPHSmooth<ValType, Ndims>::fetchNeighbours(
-      const typename SPHTree::coords &c, uint32_t newNngb, SPHState *state) {
-    ComputePrecision d2, max_dist = 0;
-    uint32_t requested = newNngb;
+template<typename ValType, int Ndims>
+void
+SPHSmooth<ValType,Ndims>::fetchNeighbours(const typename SPHTree::coords& c, uint32_t newNngb)
+{
+  ComputePrecision d2, max_dist = 0;
+  uint32_t requested = newNngb;

-    if (state != 0) {
-      state->distances = boost::shared_ptr<CoordType[]>(new CoordType[newNngb]);
-      state->ngb = boost::shared_ptr<SPHCell *[]>(new SPHCell *[newNngb]);
-    } else {
-      state = &internal;
-      if (requested > maxNgb) {
-        maxNgb = requested;
-        internal.ngb = boost::shared_ptr<P_SPHCell[]>(new P_SPHCell[maxNgb]);
-        internal.distances =
-            boost::shared_ptr<CoordType[]>(new CoordType[maxNgb]);
-      }
+  if (requested > maxNgb)
+    {
+      maxNgb = requested;
+      internal.ngb = boost::shared_ptr<P_SPHCell[]>(new P_SPHCell[maxNgb]);
+      internal.distances = boost::shared_ptr<CoordType[]>(new CoordType[maxNgb]);
    }

-    memcpy(state->currentCenter, c, sizeof(c));
-    tree->getNearestNeighbours(
-        c, requested, (SPHCell **)state->ngb.get(),
-        (CoordType *)state->distances.get());
+  memcpy(internal.currentCenter, c, sizeof(c));
+  tree->getNearestNeighbours(c, requested, (SPHCell **)internal.ngb.get(), (CoordType*)internal.distances.get());

-    state->currentNgb = 0;
-    for (uint32_t i = 0; i < requested && (state->ngb)[i] != 0;
-         i++, state->currentNgb++) {
-      state->distances[i] = sqrt(state->distances[i]);
-      d2 = state->distances[i];
+  internal.currentNgb = 0;
+  for (uint32_t i = 0; i < requested && (internal.ngb)[i] != 0; i++,internal.currentNgb++)
+    {
+      internal.distances[i] = sqrt(internal.distances[i]);
+      d2 = internal.distances[i];
      if (d2 > max_dist)
        max_dist = d2;
    }

-    state->smoothRadius = max_dist / 2;
-  }
+  internal.smoothRadius = max_dist / 2;  
+}

-  template <typename ValType, int Ndims>
-  void SPHSmooth<ValType, Ndims>::fetchNeighbours(
-      const typename SPHTree::coords &c, SPHState *state) {
-    ComputePrecision d2, max_dist = 0;
-    uint32_t requested = Nsph;
+template<typename ValType, int Ndims>
+void SPHSmooth<ValType,Ndims>::fetchNeighbours(const typename SPHTree::coords& c, SPHState *state)
+{
+  ComputePrecision d2, max_dist = 0;
+  uint32_t requested = Nsph;

-    if (state != 0) {
-      state->distances = boost::shared_ptr<CoordType[]>(new CoordType[Nsph]);
-      state->ngb = boost::shared_ptr<SPHCell *[]>(new SPHCell *[Nsph]);
-    } else
-      state = &internal;
+  if (state != 0) {
+    state->distances = boost::shared_ptr<CoordType[]>(new CoordType[Nsph]);
+    state->ngb = boost::shared_ptr<SPHCell *[]>(new SPHCell *[Nsph]);
+  } else
+    state = &internal;
    
-    memcpy(state->currentCenter, c, sizeof(c));
+  memcpy(state->currentCenter, c, sizeof(c));
  
-    tree->getNearestNeighbours(
-        c, requested, state->ngb.get(), state->distances.get());
+  tree->getNearestNeighbours(c, requested, state->ngb.get(), state->distances.get());

-    state->currentNgb = 0;
-    for (uint32_t i = 0; i < requested && state->ngb[i] != 0;
-         i++, state->currentNgb++) {
-      d2 = state->distances[i] = sqrt(state->distances[i]);
-      if (d2 > max_dist)
-        max_dist = d2;
-    }
-
-    state->smoothRadius = max_dist / 2;
-  }
-
-  template <typename ValType, int Ndims>
-  void SPHSmooth<ValType, Ndims>::fetchNeighboursOnVolume(
-      const typename SPHTree::coords &c, ComputePrecision radius) {
-    uint32_t numPart;
-    ComputePrecision d2, max_dist = 0;
-
-    memcpy(internal.currentCenter, c, sizeof(c));
-
-    internal.currentNgb = tree->getIntersection(
-        c, radius, internal.ngb, internal.distances, maxNgb);
-
-    for (uint32_t i = 0; i < internal.currentNgb; i++) {
+  state->currentNgb = 0;
+  for (uint32_t i = 0; i < requested && state->ngb[i] != 0; i++,state->currentNgb++)
+    {
      d2 = internal.distances[i] = sqrt(internal.distances[i]);
      if (d2 > max_dist)
        max_dist = d2;
    }
-    internal.smoothRadius = max_dist / 2;
-  }

-  template <typename ValType, int Ndims>
-  template <typename FuncT>
-  ComputePrecision SPHSmooth<ValType, Ndims>::computeSmoothedValue(
-      const typename SPHTree::coords &c, FuncT fun, SPHState *state) {
-    if (state == 0)
-      state = &internal;
+  state->smoothRadius = max_dist / 2;  
+}

-    ComputePrecision outputValue = 0;
-    ComputePrecision max_dist = 0;
-    ComputePrecision r3 = cube(state->smoothRadius);

-    for (uint32_t i = 0; i < state->currentNgb; i++) {
-      outputValue +=
-          computeWValue(c, *state->ngb[i], state->distances[i], fun, state);
+template<typename ValType, int Ndims>
+void
+SPHSmooth<ValType,Ndims>::fetchNeighboursOnVolume(const typename SPHTree::coords& c,
+                                                  ComputePrecision radius)
+{
+  uint32_t numPart;
+  ComputePrecision d2, max_dist = 0;
+
+  memcpy(internal.currentCenter, c, sizeof(c));
+
+  internal.currentNgb = tree->getIntersection(c, radius, internal.ngb, internal.distances,
+				  maxNgb);
+
+  for (uint32_t i = 0; i < internal.currentNgb; i++)
+    {
+      d2 = internal.distances[i] = sqrt(internal.distances[i]);
+      if (d2 > max_dist)
+        max_dist = d2;
+    }
+  internal.smoothRadius = max_dist / 2;  
+}				
+
+template<typename ValType, int Ndims>
+template<typename FuncT>
+ComputePrecision 
+SPHSmooth<ValType,Ndims>::computeSmoothedValue(const typename SPHTree::coords& c,
+                                               FuncT fun, SPHState *state)
+{
+  if (state == 0)
+    state = &internal;
+    
+  ComputePrecision outputValue = 0;
+  ComputePrecision max_dist = 0;
+  ComputePrecision r3 = cube(state->smoothRadius);
+
+  for (uint32_t i = 0; i < state->currentNgb; i++)
+    {
+      outputValue += computeWValue(c, *state->ngb[i], state->distances[i], fun, state);
    }

-    return outputValue / r3;
-  }
+  return outputValue / r3;
+}

-  template <typename ValType>
-  ComputePrecision interpolateOne(const ValType &t) {
-    return 1.0;
-  }
+template<typename ValType>
+ComputePrecision interpolateOne(const ValType& t)
+{
+  return 1.0;
+}

-  template <typename ValType, int Ndims>
-  template <typename FuncT>
-  void SPHSmooth<ValType, Ndims>::computeAdjointGradientSmoothedValue(
-      const typename SPHTree::coords &c, ComputePrecision ag_value, FuncT fun,
-      SPHState *state) {
-    if (state == 0)
-      state = &internal;
+// WARNING ! Cell's weight must be 1 !!!
+template<typename ValType, int Ndims>
+template<typename FuncT>
+ComputePrecision SPHSmooth<ValType,Ndims>::computeInterpolatedValue(const typename SPHTree::coords& c,
+								    FuncT fun, SPHState *state)
+{
+  if (state == 0)
+    state = &internal;
    
-    ComputePrecision outputValue = 0;
-    ComputePrecision max_dist = 0;
-    ComputePrecision weight = 0;
+  ComputePrecision outputValue = 0;
+  ComputePrecision max_dist = 0;
+  ComputePrecision weight = 0;

-    for (uint32_t i = 0; i < state->currentNgb; i++) {
-      weight +=
-          computeWValue(c, *state->ngb[i], state->distances[i], interpolateOne);
-    }
-
-    for (uint32_t i = 0; i < state->currentNgb; i++) {
-      auto &cell = *state->ngb[i];
-      double partial_ag =
-          computeWValue(
-              c, cell, state->distances[i],
-              [ag_value](ComputePrecision) { return ag_value; }) /
-          weight;
-      fun(cell.val.pValue, ag_value);
-    }
-  }
-
-  // WARNING ! Cell's weight must be 1 !!!
-  template <typename ValType, int Ndims>
-  template <typename FuncT>
-  ComputePrecision SPHSmooth<ValType, Ndims>::computeInterpolatedValue(
-      const typename SPHTree::coords &c, FuncT fun, SPHState *state) {
-    if (state == 0)
-      state = &internal;
-
-    ComputePrecision outputValue = 0;
-    ComputePrecision max_dist = 0;
-    ComputePrecision weight = 0;
-
-    for (uint32_t i = 0; i < state->currentNgb; i++) {
+  for (uint32_t i = 0; i < state->currentNgb; i++)
+    {
      outputValue += computeWValue(c, *state->ngb[i], state->distances[i], fun);
-      weight +=
-          computeWValue(c, *state->ngb[i], state->distances[i], interpolateOne);
+      weight += computeWValue(c, *state->ngb[i], state->distances[i], interpolateOne);
    }

-    return (outputValue == 0) ? 0 : (outputValue / weight);
-  }
+  return (outputValue == 0) ? 0 : (outputValue / weight);  
+}

-  template <typename ValType, int Ndims>
-  template <typename FuncT>
-  void
-  SPHSmooth<ValType, Ndims>::runForEachNeighbour(FuncT fun, SPHState *state) {
-    if (state == 0)
-      state = &internal;
+template<typename ValType, int Ndims>
+template<typename FuncT>
+void SPHSmooth<ValType,Ndims>::runForEachNeighbour(FuncT fun, SPHState *state)
+{
+  if (state == 0)
+    state = &internal;
    
-    for (uint32_t i = 0; i < state->currentNgb; i++) {
+  for (uint32_t i = 0; i < state->currentNgb; i++)
+    {
      fun(state->ngb[i]);
    }
-  }
+}

-  template <typename ValType, int Ndims>
-  void SPHSmooth<ValType, Ndims>::addGridSite(
-      const typename SPHTree::coords &c, SPHState *state) {
-    ComputePrecision outputValue = 0;
-    ComputePrecision max_dist = 0;
-    ComputePrecision r3 = cube(state->smoothRadius);

-    for (uint32_t i = 0; i < state->currentNgb; i++) {
-      ComputePrecision d = state->distances[i];
-      SPHCell &cell = *(state->ngb[i]);
-      double kernel_value = getKernel(d / state->smoothRadius) / r3;
-#pragma omp atomic update
-      cell.val.weight += kernel_value;
+template<typename ValType, int Ndims>
+void SPHSmooth<ValType,Ndims>::addGridSite(const typename SPHTree::coords& c)
+{
+  ComputePrecision outputValue = 0;
+  ComputePrecision max_dist = 0;
+  
+  ComputePrecision r3 = cube(internal.smoothRadius);
+
+  for (uint32_t i = 0; i < internal.currentNgb; i++)
+    {
+      ComputePrecision d = internal.distances[i];
+      SPHCell& cell = *(internal.ngb[i]);
+      cell.val.weight += getKernel(d/internal.smoothRadius) / r3;
    }
-  }
+}

-  template <typename ValType, int Ndims>
-  void
-  SPHSmooth<ValType, Ndims>::addGridSite(const typename SPHTree::coords &c) {
-    addGridSite(c, &internal);
-  }
-
-  template <typename ValType, int Ndims>
-  ComputePrecision
-  SPHSmooth<ValType, Ndims>::getKernel(ComputePrecision x) const {
-    // WARNING !!! This is an unnormalized version of the kernel.
-    if (x < 1)
-      return 1 - 1.5 * x * x + 0.75 * x * x * x;
-    else if (x < 2) {
+template<typename ValType, int Ndims>
+ComputePrecision 
+SPHSmooth<ValType,Ndims>::getKernel(ComputePrecision x) const
+{
+  // WARNING !!! This is an unnormalized version of the kernel.
+  if (x < 1)
+    return 1 - 1.5 * x * x + 0.75 * x * x * x;
+  else if (x < 2)
+    {
      ComputePrecision d = 2 - x;
      return 0.25 * d * d * d;
-    } else
-      return 0;
-  }
+    }
+  else
+    return 0;
+}

-  template <typename ValType, int Ndims>
-  bool SPHSmooth<ValType, Ndims>::hasNeighbours() const {
-    return (internal.currentNgb != 0);
-  }
+template<typename ValType, int Ndims>
+bool SPHSmooth<ValType,Ndims>::hasNeighbours() const
+{
+  return (internal.currentNgb != 0);
+}

-  template <class ValType1, class ValType2, int Ndims>
-  bool operator<(
-      const SPHSmooth<ValType1, Ndims> &s1,
-      const SPHSmooth<ValType2, Ndims> &s2) {
-    return (s1.getSmoothingLen() < s2.getSmoothingLen());
-  }
-}; // namespace CosmoTool
+template<class ValType1, class ValType2, int Ndims>
+bool operator<(const SPHSmooth<ValType1, Ndims>& s1, const SPHSmooth<ValType2, Ndims>& s2)
+{
+  return (s1.getSmoothingLen() < s2.getSmoothingLen());
+}
+
+};
--- a/src/yorick_nc3.cpp
+++ b/src/yorick_nc3.cpp
@ -260,6 +260,7 @@ namespace CosmoTool {
  template<typename T>
  void loadArray(const std::string& fname,
 		 T*&array, uint32_t *&dimList, uint32_t& rank)
+	throw (NoSuchFileException)
  {
    NcFile f(fname.c_str(), NcFile::ReadOnly);