GitLab

This should be the total number not the local one. We might be in parallel and a core could have less than 3 particles.

Ippl reduce

See merge request !23

src/CMakeLists.txt: target_link_libraries( OPALib ...) added again, don't understand why this is required

- use CMake's 'MPI_CXX_LIBRARIES'
- reviewed

move SDDSParser from OPAL to Opt-Pilot and OpalSimulation from Opt-Pilot to OPAL to avoid a dependence of Opt-Pilot from OPAL

Had to add the average longitudinal momentum to the momentum individual to the particles as with other injected distributions. Fixes  #194

- add "allreduce" and "reduce" functions
- remove unused variable "minLocNum_m"
- use new communication methods in PartBunchBase<T, Dim>::calcMoments()
  and PartBunchBase<T, Dim>::calcBeamParameters()

modified:   CMakeLists.txt
modified:   GlobalComm.h
modified:   GlobalComm.hpp
new file:   Operations.h
modified:   ../../../src/Classic/Algorithms/PartBunchBase.h
modified:   ../../../src/Classic/Algorithms/PartBunchBase.hpp

- replace "allreduce" with gather.
- remove partPerNode_m

new file:   DataTypes.h
modified:   GlobalComm.h
modified:   GlobalComm.hpp
modified:   ../../../src/Classic/Algorithms/PartBunchBase.h
modified:   ../../../src/Classic/Algorithms/PartBunchBase.hpp

- change all 'unsigned int' to 'site_t'
- length of sequence is 'size_t' not ptr to 'int'

- revied
- libgfortran removed for psa-standalone.exe

- out-commented enable Fortran removed

- reformat with indent equal 4, split lines to make code better readable
- reviewed