Pushing everything again, accidentally deleted my remote repository
This commit is contained in:
jakob.schratter
commit
1bee3e8e5b
101 changed files with 9428 additions and 0 deletions
123
ex1/CLANG_default.mk
Normal file
123
ex1/CLANG_default.mk
Normal file
|
|
@ -0,0 +1,123 @@
|
|||
# Basic Defintions for using GNU-compiler suite sequentially
|
||||
# requires setting of COMPILER=CLANG_
|
||||
|
||||
#CLANGPATH=//usr/lib/llvm-10/bin/
|
||||
CC = ${CLANGPATH}clang
|
||||
CXX = ${CLANGPATH}clang++
|
||||
#CXX = ${CLANGPATH}clang++ -lomptarget -fopenmp-targets=nvptx64-nvidia-cuda --cuda-path=/opt/pgi/linux86-64/2017/cuda/8.0
|
||||
#F77 = gfortran
|
||||
LINKER = ${CXX}
|
||||
|
||||
#http://clang.llvm.org/docs/UsersManual.html#options-to-control-error-and-warning-messages
|
||||
WARNINGS += -Weverything -Wno-c++98-compat -Wno-sign-conversion -Wno-date-time -Wno-shorten-64-to-32 -Wno-padded -ferror-limit=1
|
||||
WARNINGS += -Wdocumentation -Wconversion -Wshadow -Wfloat-conversion -pedantic
|
||||
#-fsyntax-only -Wdocumentation -Wconversion -Wshadow -Wfloat-conversion -pedantic
|
||||
|
||||
CXXFLAGS += -O3 -std=c++17 -ferror-limit=1 ${WARNINGS}
|
||||
# don't use -Ofast
|
||||
# -ftrapv
|
||||
LINKFLAGS += -O3
|
||||
|
||||
# different libraries in Ubuntu or manajaró
|
||||
ifndef UBUNTU
|
||||
UBUNTU=1
|
||||
endif
|
||||
|
||||
# BLAS, LAPACK
|
||||
LINKFLAGS += -llapack -lblas
|
||||
# -lopenblas
|
||||
ifeq ($(UBUNTU),1)
|
||||
# ubuntu
|
||||
else
|
||||
# on archlinux
|
||||
LINKFLAGS += -lcblas
|
||||
endif
|
||||
|
||||
# interprocedural optimization
|
||||
CXXFLAGS += -flto
|
||||
LINKFLAGS += -flto
|
||||
|
||||
# very good check
|
||||
# http://clang.llvm.org/extra/clang-tidy/
|
||||
# good check, see: http://llvm.org/docs/CodingStandards.html#include-style
|
||||
SWITCH_OFF=,-readability-magic-numbers,-readability-redundant-control-flow,-readability-redundant-member-init
|
||||
SWITCH_OFF+=,-readability-redundant-member-init,-readability-isolate-declaration
|
||||
#READABILITY=,readability*${SWITCH_OFF}
|
||||
#TIDYFLAGS = -checks=llvm-*,-llvm-header-guard -header-filter=.* -enable-check-profile -extra-arg="-std=c++17" -extra-arg="-fopenmp"
|
||||
TIDYFLAGS = -checks=llvm-*,-llvm-header-guard${READABILITY} -header-filter=.* -enable-check-profile -extra-arg="-std=c++17" -extra-arg="-fopenmp"
|
||||
#TIDYFLAGS += -checks='modernize*
|
||||
# ???
|
||||
#TIDYFLAGS = -checks='cert*' -header-filter=.*
|
||||
# MPI checks ??
|
||||
#TIDYFLAGS = -checks='mpi*'
|
||||
# ??
|
||||
#TIDYFLAGS = -checks='performance*' -header-filter=.*
|
||||
#TIDYFLAGS = -checks='portability-*' -header-filter=.*
|
||||
#TIDYFLAGS = -checks='readability-*' -header-filter=.*
|
||||
|
||||
default: ${PROGRAM}
|
||||
|
||||
${PROGRAM}: ${OBJECTS}
|
||||
$(LINKER) $^ ${LINKFLAGS} -o $@
|
||||
|
||||
clean:
|
||||
@rm -f ${PROGRAM} ${OBJECTS}
|
||||
|
||||
clean_all:: clean
|
||||
@rm -f *_ *~ *.bak *.log *.out *.tar
|
||||
|
||||
codecheck: tidy_check
|
||||
tidy_check:
|
||||
clang-tidy ${SOURCES} ${TIDYFLAGS} -- ${SOURCES}
|
||||
# see also http://clang-developers.42468.n3.nabble.com/Error-while-trying-to-load-a-compilation-database-td4049722.html
|
||||
|
||||
run: clean ${PROGRAM}
|
||||
# time ./${PROGRAM} ${PARAMS}
|
||||
./${PROGRAM} ${PARAMS}
|
||||
|
||||
# tar the current directory
|
||||
MY_DIR = `basename ${PWD}`
|
||||
tar: clean_all
|
||||
@echo "Tar the directory: " ${MY_DIR}
|
||||
@cd .. ;\
|
||||
tar cf ${MY_DIR}.tar ${MY_DIR} *default.mk ;\
|
||||
cd ${MY_DIR}
|
||||
# tar cf `basename ${PWD}`.tar *
|
||||
|
||||
doc:
|
||||
doxygen Doxyfile
|
||||
|
||||
#########################################################################
|
||||
|
||||
.cpp.o:
|
||||
$(CXX) -c $(CXXFLAGS) -o $@ $<
|
||||
|
||||
.c.o:
|
||||
$(CC) -c $(CFLAGS) -o $@ $<
|
||||
|
||||
.f.o:
|
||||
$(F77) -c $(FFLAGS) -o $@ $<
|
||||
|
||||
##################################################################################################
|
||||
# some tools
|
||||
# Cache behaviour (CXXFLAGS += -g tracks down to source lines; no -pg in linkflags)
|
||||
cache: ${PROGRAM}
|
||||
valgrind --tool=callgrind --simulate-cache=yes ./$^ ${PARAMS}
|
||||
# kcachegrind callgrind.out.<pid> &
|
||||
kcachegrind `ls -1tr callgrind.out.* |tail -1`
|
||||
|
||||
# Check for wrong memory accesses, memory leaks, ...
|
||||
# use smaller data sets
|
||||
mem: ${PROGRAM}
|
||||
valgrind -v --leak-check=yes --tool=memcheck --undef-value-errors=yes --track-origins=yes --log-file=$^.addr.out --show-reachable=yes ./$^ ${PARAMS}
|
||||
|
||||
# Simple run time profiling of your code
|
||||
# CXXFLAGS += -g -pg
|
||||
# LINKFLAGS += -pg
|
||||
prof: ${PROGRAM}
|
||||
perf record ./$^ ${PARAMS}
|
||||
perf report
|
||||
# gprof -b ./$^ > gp.out
|
||||
# kprof -f gp.out -p gprof &
|
||||
|
||||
codecheck: tidy_check
|
||||
130
ex1/GCC_AMD32_default.mk
Normal file
130
ex1/GCC_AMD32_default.mk
Normal file
|
|
@ -0,0 +1,130 @@
|
|||
# Basic Defintions for using GNU-compiler suite sequentially
|
||||
# requires setting of COMPILER=GCC_
|
||||
|
||||
CC = gcc
|
||||
CXX = g++
|
||||
F77 = gfortran
|
||||
LINKER = ${CXX}
|
||||
|
||||
# on mephisto:
|
||||
#CXXFLAGS += -I/share/apps/atlas/include
|
||||
#LINKFLAGS += -L/share/apps/atlas/lib
|
||||
#LINKFLAGS += -lcblas -latlas
|
||||
|
||||
#LINKFLAGS += -lblas
|
||||
# Der <cblas.h> Header muss mit extern "C" versehen werden, damit g++ alles findet.
|
||||
|
||||
|
||||
#WARNINGS = -pedantic -pedantic-errors -Wall -Wextra -Werror -Wconversion -Weffc++ -Woverloaded-virtual -Wfloat-equal -Wshadow
|
||||
WARNINGS = -pedantic -Wall -Wextra -Wconversion -Weffc++ -Woverloaded-virtual -Wfloat-equal -Wshadow \
|
||||
-Wredundant-decls -Winline -fmax-errors=1
|
||||
# -Wunreachable-code
|
||||
# -Wunreachable-code
|
||||
CXXFLAGS += -ffast-math -O3 -march=native -std=c++17 ${WARNINGS}
|
||||
#CXXFLAGS += -Ofast -funroll-all-loops -std=c++17 ${WARNINGS}
|
||||
#-msse3
|
||||
# -ftree-vectorizer-verbose=2 -DNDEBUG
|
||||
# -ftree-vectorizer-verbose=5
|
||||
# -ftree-vectorize -fdump-tree-vect-blocks=foo.dump -fdump-tree-pre=stderr
|
||||
|
||||
# CFLAGS = -ffast-math -O3 -DNDEBUG -msse3 -fopenmp -fdump-tree-vect-details
|
||||
# CFLAGS = -ffast-math -O3 -funroll-loops -DNDEBUG -msse3 -fopenmp -ftree-vectorizer-verbose=2
|
||||
# #CFLAGS = -ffast-math -O3 -DNDEBUG -msse3 -fopenmp
|
||||
# FFLAGS = -ffast-math -O3 -DNDEBUG -msse3 -fopenmp
|
||||
# LFLAGS = -ffast-math -O3 -DNDEBUG -msse3 -fopenmp
|
||||
LINKFLAGS += -O3
|
||||
|
||||
# BLAS, LAPACK
|
||||
OPENBLAS_DIR = /opt/openblas_GCCseq
|
||||
#OPENBLAS_DIR = /opt/openblas_GCC
|
||||
OPENBLAS_LIBDIR = ${OPENBLAS_DIR}/lib
|
||||
OPENBLAS_INCDIR = ${OPENBLAS_DIR}/include
|
||||
CXXFLAGS += -I${OPENBLAS_INCDIR}
|
||||
LINKFLAGS += -L${OPENBLAS_LIBDIR} -lopenblas
|
||||
|
||||
# interprocedural optimization
|
||||
CXXFLAGS += -flto
|
||||
LINKFLAGS += -flto
|
||||
|
||||
# profiling tools
|
||||
#CXXFLAGS += -pg
|
||||
#LINKFLAGS += -pg
|
||||
|
||||
default: ${PROGRAM}
|
||||
|
||||
${PROGRAM}: ${OBJECTS}
|
||||
$(LINKER) $^ ${LINKFLAGS} -o $@
|
||||
|
||||
clean:
|
||||
@rm -f ${PROGRAM} ${OBJECTS}
|
||||
|
||||
clean_all:: clean
|
||||
-@rm -f *_ *~ *.bak *.log *.out *.tar *.orig
|
||||
-@rm -r html
|
||||
|
||||
run: clean ${PROGRAM}
|
||||
# time ./${PROGRAM}
|
||||
# ./${PROGRAM}
|
||||
( export LD_LIBRARY_PATH=${OPENBLAS_LIBDIR}:${LD_LIBRARY_PATH} ; ./${PROGRAM} )
|
||||
# or 'export LD_LIBRARY_PATH=/opt/openblas_gcc/lib:${LD_LIBRARY_PATH}' in your ~/.bashrc
|
||||
|
||||
# tar the current directory
|
||||
MY_DIR = `basename ${PWD}`
|
||||
tar:
|
||||
@echo "Tar the directory: " ${MY_DIR}
|
||||
@cd .. ;\
|
||||
tar cf ${MY_DIR}.tar ${MY_DIR} *default.mk ;\
|
||||
cd ${MY_DIR}
|
||||
# tar cf `basename ${PWD}`.tar *
|
||||
|
||||
doc:
|
||||
doxygen Doxyfile
|
||||
|
||||
#########################################################################
|
||||
|
||||
.cpp.o:
|
||||
$(CXX) -c $(CXXFLAGS) -o $@ $<
|
||||
|
||||
.c.o:
|
||||
$(CC) -c $(CFLAGS) -o $@ $<
|
||||
|
||||
.f.o:
|
||||
$(F77) -c $(FFLAGS) -o $@ $<
|
||||
|
||||
##################################################################################################
|
||||
# some tools
|
||||
# Cache behaviour (CXXFLAGS += -g tracks down to source lines; no -pg in linkflags)
|
||||
cache: ${PROGRAM}
|
||||
valgrind --tool=callgrind --simulate-cache=yes ./$^
|
||||
# kcachegrind callgrind.out.<pid> &
|
||||
kcachegrind `ls -1tr callgrind.out.* |tail -1`
|
||||
|
||||
# Check for wrong memory accesses, memory leaks, ...
|
||||
# use smaller data sets
|
||||
# no "-pg" in compile/link options
|
||||
mem: ${PROGRAM}
|
||||
valgrind -v --leak-check=yes --tool=memcheck --undef-value-errors=yes --track-origins=yes --log-file=$^.addr.out --show-reachable=yes ./$^
|
||||
|
||||
# Simple run time profiling of your code
|
||||
# CXXFLAGS += -g -pg
|
||||
# LINKFLAGS += -pg
|
||||
prof: ${PROGRAM}
|
||||
./$^
|
||||
gprof -b ./$^ > gp.out
|
||||
# kprof -f gp.out -p gprof &
|
||||
|
||||
#Trace your heap:
|
||||
#> heaptrack ./main.GCC_
|
||||
#> heaptrack_gui heaptrack.main.GCC_.<pid>.gz
|
||||
heap: ${PROGRAM}
|
||||
heaptrack ./$^ 11
|
||||
heaptrack_gui `ls -1tr heaptrack.$^.* |tail -1` &
|
||||
|
||||
|
||||
|
||||
########################################################################
|
||||
# get the detailed status of all optimization flags
|
||||
info:
|
||||
echo "detailed status of all optimization flags"
|
||||
$(CXX) --version
|
||||
$(CXX) -Q $(CXXFLAGS) --help=optimizers
|
||||
183
ex1/GCC_default.mk
Normal file
183
ex1/GCC_default.mk
Normal file
|
|
@ -0,0 +1,183 @@
|
|||
# Basic Defintions for using GNU-compiler suite sequentially
|
||||
# requires setting of COMPILER=GCC_
|
||||
|
||||
CC = gcc
|
||||
CXX = g++
|
||||
F77 = gfortran
|
||||
LINKER = ${CXX}
|
||||
|
||||
WARNINGS = -Wall -pedantic -Wextra -Weffc++ -Woverloaded-virtual -Wfloat-equal -Wshadow \
|
||||
-Wredundant-decls -Winline -fmax-errors=1
|
||||
# -Wunreachable-code
|
||||
CXXFLAGS += -ffast-math -O3 -march=native -std=c++17 ${WARNINGS}
|
||||
#CXXFLAGS += -Ofast -funroll-all-loops -std=c++17 ${WARNINGS}
|
||||
#-msse3
|
||||
# -ftree-vectorizer-verbose=2 -DNDEBUG
|
||||
# -ftree-vectorizer-verbose=5
|
||||
# -ftree-vectorize -fdump-tree-vect-blocks=foo.dump -fdump-tree-pre=stderr
|
||||
|
||||
# CFLAGS = -ffast-math -O3 -DNDEBUG -msse3 -fopenmp -fdump-tree-vect-details
|
||||
# CFLAGS = -ffast-math -O3 -funroll-loops -DNDEBUG -msse3 -fopenmp -ftree-vectorizer-verbose=2
|
||||
# #CFLAGS = -ffast-math -O3 -DNDEBUG -msse3 -fopenmp
|
||||
# FFLAGS = -ffast-math -O3 -DNDEBUG -msse3 -fopenmp
|
||||
# LFLAGS = -ffast-math -O3 -DNDEBUG -msse3 -fopenmp
|
||||
LINKFLAGS += -O3
|
||||
|
||||
#architecture
|
||||
#CPU = -march=znver2
|
||||
CXXFLAGS += ${CPU}
|
||||
LINKFLAGS += ${CPU}
|
||||
|
||||
# different libraries in Ubuntu or manajaró
|
||||
ifndef UBUNTU
|
||||
UBUNTU=1
|
||||
endif
|
||||
|
||||
# BLAS, LAPACK
|
||||
ifeq ($(UBUNTU),1)
|
||||
LINKFLAGS += -llapack -lblas
|
||||
# -lopenblas
|
||||
else
|
||||
# on archlinux
|
||||
LINKFLAGS += -llapack -lopenblas -lcblas
|
||||
endif
|
||||
|
||||
# interprocedural optimization
|
||||
CXXFLAGS += -flto
|
||||
LINKFLAGS += -flto
|
||||
|
||||
# for debugging purpose (save code)
|
||||
# -fsanitize=leak # only one out the three can be used
|
||||
# -fsanitize=address
|
||||
# -fsanitize=thread
|
||||
SANITARY = -fsanitize=address -fsanitize=undefined -fsanitize=null -fsanitize=return \
|
||||
-fsanitize=bounds -fsanitize=alignment -fsanitize=float-divide-by-zero -fsanitize=float-cast-overflow \
|
||||
-fsanitize=bool -fsanitize=enum -fsanitize=vptr
|
||||
#CXXFLAGS += ${SANITARY}
|
||||
#LINKFLAGS += ${SANITARY}
|
||||
|
||||
# profiling tools
|
||||
#CXXFLAGS += -pg
|
||||
#LINKFLAGS += -pg
|
||||
|
||||
|
||||
default: ${PROGRAM}
|
||||
|
||||
${PROGRAM}: ${OBJECTS}
|
||||
$(LINKER) $^ ${LINKFLAGS} -o $@
|
||||
|
||||
clean:
|
||||
@rm -f ${PROGRAM} ${OBJECTS}
|
||||
|
||||
clean_all:: clean
|
||||
-@rm -f *_ *~ *.bak *.log *.out *.tar *.orig *.optrpt
|
||||
-@rm -rf html
|
||||
|
||||
run: clean ${PROGRAM}
|
||||
#run: ${PROGRAM}
|
||||
# time ./${PROGRAM} ${PARAMS}
|
||||
./${PROGRAM} ${PARAMS}
|
||||
|
||||
# tar the current directory
|
||||
MY_DIR = `basename ${PWD}`
|
||||
tar: clean_all
|
||||
@echo "Tar the directory: " ${MY_DIR}
|
||||
@cd .. ;\
|
||||
tar cf ${MY_DIR}.tar ${MY_DIR} *default.mk ;\
|
||||
cd ${MY_DIR}
|
||||
# tar cf `basename ${PWD}`.tar *
|
||||
#find . -size +10M > large_files
|
||||
#--exclude-from ${MY_DIR}/large_files
|
||||
|
||||
zip: clean
|
||||
@echo "Zip the directory: " ${MY_DIR}
|
||||
@cd .. ;\
|
||||
zip -r ${MY_DIR}.zip ${MY_DIR} *default.mk ;\
|
||||
cd ${MY_DIR}
|
||||
|
||||
doc:
|
||||
doxygen Doxyfile
|
||||
|
||||
#########################################################################
|
||||
.SUFFIXES: .f90
|
||||
|
||||
.cpp.o:
|
||||
$(CXX) -c $(CXXFLAGS) -o $@ $<
|
||||
# $(CXX) -c $(CXXFLAGS) -o $@ $< 2>&1 | tee -a $<.log
|
||||
# $(CXX) -c $(CXXFLAGS) -o $@ $< 2>&1 | tee -a $(<:.cpp=.log)
|
||||
|
||||
.c.o:
|
||||
$(CC) -c $(CFLAGS) -o $@ $<
|
||||
|
||||
.f.o:
|
||||
$(F77) -c $(FFLAGS) -o $@ $<
|
||||
|
||||
.f90.o:
|
||||
$(F77) -c $(FFLAGS) -o $@ $<
|
||||
|
||||
##################################################################################################
|
||||
# some tools
|
||||
# Cache behaviour (CXXFLAGS += -g tracks down to source lines; no -pg in linkflags)
|
||||
cache: ${PROGRAM}
|
||||
valgrind --tool=callgrind --simulate-cache=yes ./$^ ${PARAMS}
|
||||
# kcachegrind callgrind.out.<pid> &
|
||||
kcachegrind `ls -1tr callgrind.out.* |tail -1`
|
||||
|
||||
# Check for wrong memory accesses, memory leaks, ...
|
||||
# use smaller data sets
|
||||
# no "-pg" in compile/link options
|
||||
mem: ${PROGRAM}
|
||||
valgrind -v --leak-check=yes --tool=memcheck --undef-value-errors=yes --track-origins=yes --log-file=$^.addr.out --show-reachable=yes ./$^ ${PARAMS}
|
||||
# Graphical interface
|
||||
# valkyrie
|
||||
|
||||
# Simple run time profiling of your code
|
||||
# CXXFLAGS += -g -pg
|
||||
# LINKFLAGS += -pg
|
||||
prof: ${PROGRAM}
|
||||
perf record ./$^ ${PARAMS}
|
||||
perf report
|
||||
# gprof -b ./$^ > gp.out
|
||||
# kprof -f gp.out -p gprof &
|
||||
|
||||
# perf in Ubuntu 20.04: https://www.howtoforge.com/how-to-install-perf-performance-analysis-tool-on-ubuntu-20-04/
|
||||
# * install
|
||||
# * sudo vi /etc/sysctl.conf
|
||||
# add kernel.perf_event_paranoid = 0
|
||||
|
||||
#Trace your heap:
|
||||
#> heaptrack ./main.GCC_
|
||||
#> heaptrack_gui heaptrack.main.GCC_.<pid>.gz
|
||||
heap: ${PROGRAM}
|
||||
heaptrack ./$^ ${PARAMS}
|
||||
heaptrack_gui `ls -1tr heaptrack.$^.* |tail -1` &
|
||||
|
||||
codecheck: $(SOURCES)
|
||||
cppcheck --enable=all --inconclusive --std=c++17 --suppress=missingIncludeSystem $^
|
||||
|
||||
|
||||
########################################################################
|
||||
# get the detailed status of all optimization flags
|
||||
info:
|
||||
echo "detailed status of all optimization flags"
|
||||
$(CXX) --version
|
||||
$(CXX) -Q $(CXXFLAGS) --help=optimizers
|
||||
lscpu
|
||||
inxi -C
|
||||
lstopo
|
||||
|
||||
# Excellent hardware info
|
||||
# hardinfo
|
||||
# Life monitoring of CPU frequency etc.
|
||||
# sudo i7z
|
||||
|
||||
# Memory consumption
|
||||
# vmstat -at -SM 3
|
||||
# xfce4-taskmanager
|
||||
|
||||
|
||||
# https://www.tecmint.com/check-linux-cpu-information/
|
||||
#https://www.tecmint.com/monitor-cpu-and-gpu-temperature-in-ubuntu/
|
||||
|
||||
# Debugging:
|
||||
# https://wiki.archlinux.org/index.php/Debugging
|
||||
137
ex1/ICC_default.mk
Normal file
137
ex1/ICC_default.mk
Normal file
|
|
@ -0,0 +1,137 @@
|
|||
# Basic Defintions for using INTEL compiler suite sequentially
|
||||
# requires setting of COMPILER=ICC_
|
||||
|
||||
#BINDIR = /opt/intel/bin/
|
||||
|
||||
# special on my sony [GH]
|
||||
#BINDIR = /opt/save.intel/bin/
|
||||
# very special on my sony [GH]
|
||||
# FIND_LIBS = -L /opt/save.intel/composer_xe_2013.1.117/mkl/lib/intel64/libmkl_intel_lp64.so
|
||||
|
||||
# Error with g++-4.8 using icpc14.0,x:
|
||||
# find directory wherein bits/c++config.h is located
|
||||
# 'locate bits/c++config.h'
|
||||
#FOUND_CONFIG = -I/usr/include/x86_64-linux-gnu/c++/4.8
|
||||
|
||||
|
||||
CC = ${BINDIR}icc
|
||||
CXX = ${BINDIR}icpc
|
||||
F77 = ${BINDIR}ifort
|
||||
LINKER = ${CXX}
|
||||
|
||||
|
||||
WARNINGS = -Wall -Weffc++ -Woverloaded-virtual -Wfloat-equal -Wshadow -wd2015,2012 -wn3
|
||||
# -Winline -Wredundant-decls -Wunreachable-code
|
||||
CXXFLAGS += -O3 -fargument-noalias -std=c++17 -DNDEBUG ${WARNINGS} -mkl ${FOUND_CONFIG}
|
||||
# profiling tools
|
||||
#CXXFLAGS += -pg
|
||||
#LINKFLAGS += -pg
|
||||
# -vec-report=3
|
||||
# -qopt-report=5 -qopt-report-phase=vec
|
||||
# -guide -parallel
|
||||
# -guide-opts=string -guide-par[=n] -guide-vec[=n]
|
||||
# -auto-p32 -simd
|
||||
CXXFLAGS += -align
|
||||
|
||||
# use MKL by INTEL
|
||||
# https://software.intel.com/content/www/us/en/develop/documentation/mkl-linux-developer-guide/top/linking-your-application-with-the-intel-math-kernel-library/linking-quick-start/using-the-mkl-compiler-option.html
|
||||
# https://software.intel.com/content/www/us/en/develop/articles/intel-mkl-link-line-advisor.html
|
||||
# LINKFLAGS += -L${BINDIR}../composer_xe_2013.1.117/mkl/lib/intel64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread
|
||||
#LINKFLAGS += -O3 -L/opt/intel/mkl/lib -mkl
|
||||
LINKFLAGS += -O3 -mkl=sequential
|
||||
|
||||
# interprocedural optimization
|
||||
CXXFLAGS += -ipo
|
||||
LINKFLAGS += -ipo
|
||||
|
||||
# annotated assembler file
|
||||
ANNOTED = -fsource-asm -S
|
||||
|
||||
default: ${PROGRAM}
|
||||
|
||||
${PROGRAM}: ${OBJECTS}
|
||||
$(LINKER) $^ ${LINKFLAGS} -o $@
|
||||
|
||||
clean:
|
||||
rm -f ${PROGRAM} ${OBJECTS}
|
||||
|
||||
clean_all:: clean
|
||||
@rm -f *_ *~ *.bak *.log *.out *.tar
|
||||
|
||||
run: clean ${PROGRAM}
|
||||
./${PROGRAM}
|
||||
|
||||
# tar the current directory
|
||||
MY_DIR = `basename ${PWD}`
|
||||
tar: clean_all
|
||||
@echo "Tar the directory: " ${MY_DIR}
|
||||
@cd .. ;\
|
||||
tar cf ${MY_DIR}.tar ${MY_DIR} *default.mk ;\
|
||||
cd ${MY_DIR}
|
||||
# tar cf `basename ${PWD}`.tar *
|
||||
|
||||
doc:
|
||||
doxygen Doxyfile
|
||||
|
||||
#########################################################################
|
||||
|
||||
.cpp.o:
|
||||
$(CXX) -c $(CXXFLAGS) -o $@ $<
|
||||
|
||||
.c.o:
|
||||
$(CC) -c $(CFLAGS) -o $@ $<
|
||||
|
||||
.f.o:
|
||||
$(F77) -c $(FFLAGS) -o $@ $<
|
||||
|
||||
##################################################################################################
|
||||
# # some tools
|
||||
# # Cache behaviour (CXXFLAGS += -g tracks down to source lines)
|
||||
# cache: ${PROGRAM}
|
||||
# valgrind --tool=callgrind --simulate-cache=yes ./$^
|
||||
# # kcachegrind callgrind.out.<pid> &
|
||||
#
|
||||
# # Check for wrong memory accesses, memory leaks, ...
|
||||
# # use smaller data sets
|
||||
# mem: ${PROGRAM}
|
||||
# valgrind -v --leak-check=yes --tool=memcheck --undef-value-errors=yes --track-origins=yes --log-file=$^.addr.out --show-reachable=yes ./$^
|
||||
#
|
||||
# # Simple run time profiling of your code
|
||||
# # CXXFLAGS += -g -pg
|
||||
# # LINKFLAGS += -pg
|
||||
# prof: ${PROGRAM}
|
||||
# ./$^
|
||||
# gprof -b ./$^ > gp.out
|
||||
# # kprof -f gp.out -p gprof &
|
||||
#
|
||||
|
||||
|
||||
mem: inspector
|
||||
prof: amplifier
|
||||
cache: amplifier
|
||||
|
||||
gap_par_report:
|
||||
${CXX} -c -guide -parallel $(SOURCES) 2> gap.txt
|
||||
|
||||
# GUI for performance report
|
||||
amplifier: ${PROGRAM}
|
||||
echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope
|
||||
# alternatively to the solution above:
|
||||
#edit file /etc/sysctl.d/10-ptrace.conf and set variable kernel.yama.ptrace_scope variable to 0 .
|
||||
amplxe-gui &
|
||||
|
||||
# GUI for Memory and Thread analyzer (race condition)
|
||||
inspector: ${PROGRAM}
|
||||
echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope
|
||||
inspxe-gui &
|
||||
|
||||
advisor:
|
||||
echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope
|
||||
advixe-gui &
|
||||
|
||||
icc-info:
|
||||
icpc -# main.cpp
|
||||
|
||||
|
||||
|
||||
|
||||
176
ex1/ONEAPI_default.mk
Normal file
176
ex1/ONEAPI_default.mk
Normal file
|
|
@ -0,0 +1,176 @@
|
|||
# Basic Defintions for using INTEL compiler suite sequentially
|
||||
# requires setting of COMPILER=ONEAPI_
|
||||
|
||||
# https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onemkl/link-line-advisor.html
|
||||
# requires
|
||||
# source /opt/intel/oneapi/setvars.sh
|
||||
# on AMD: export MKL_DEBUG_CPU_TYPE=5
|
||||
|
||||
#BINDIR = /opt/intel/oneapi/compiler/latest/linux/bin/
|
||||
#MKL_ROOT = /opt/intel/oneapi/mkl/latest/
|
||||
#export KMP_AFFINITY=verbose,compact
|
||||
|
||||
CC = ${BINDIR}icc
|
||||
CXX = ${BINDIR}dpcpp
|
||||
F77 = ${BINDIR}ifort
|
||||
LINKER = ${CXX}
|
||||
|
||||
## Compiler flags
|
||||
WARNINGS = -Wall -Weffc++ -Woverloaded-virtual -Wfloat-equal -Wshadow -pedantic
|
||||
WARNINGS += -Wpessimizing-move -Wredundant-move
|
||||
#-wd2015,2012,2014 -wn3
|
||||
# -Winline -Wredundant-decls -Wunreachable-code
|
||||
# -qopt-subscript-in-range
|
||||
# -vec-threshold0
|
||||
|
||||
CXXFLAGS += -O3 -std=c++17 ${WARNINGS}
|
||||
#CXXFLAGS += -DMKL_ILP64 -I"${MKLROOT}/include"
|
||||
#CXXFLAGS += -DMKL_ILP32 -I"${MKLROOT}/include"
|
||||
LINKFLAGS += -O3
|
||||
|
||||
# interprocedural optimization
|
||||
CXXFLAGS += -ipo
|
||||
LINKFLAGS += -ipo
|
||||
LINKFLAGS += -flto
|
||||
|
||||
# annotated Assembler file
|
||||
ANNOTED = -fsource-asm -S
|
||||
|
||||
#architecture
|
||||
CPU = -march=core-avx2
|
||||
#CPU += -mtp=zen
|
||||
# -xCORE-AVX2
|
||||
# -axcode COMMON-AVX512 -axcode MIC-AVX512 -axcode CORE-AVX512 -axcode CORE-AVX2
|
||||
CXXFLAGS += ${CPU}
|
||||
LINKFLAGS += ${CPU}
|
||||
|
||||
# use MKL by INTEL
|
||||
# https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onemkl/link-line-advisor.html
|
||||
# sequential MKL
|
||||
# use the 32 bit interface (LP64) instead of 64 bit interface (ILP64)
|
||||
CXXFLAGS += -qmkl=sequential -UMKL_ILP64
|
||||
LINKFLAGS += -O3 -qmkl=sequential -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread
|
||||
#LINKFLAGS += -O3 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread
|
||||
|
||||
# shared libs: https://aur.archlinux.org/packages/intel-oneapi-compiler-static
|
||||
# install intel-oneapi-compiler-static
|
||||
# or
|
||||
LINKFLAGS += -shared-intel
|
||||
|
||||
|
||||
OPENMP = -qopenmp
|
||||
CXXFLAGS += ${OPENMP}
|
||||
LINKFLAGS += ${OPENMP}
|
||||
|
||||
|
||||
# profiling tools
|
||||
#CXXFLAGS += -pg
|
||||
#LINKFLAGS += -pg
|
||||
# -vec-report=3
|
||||
# -qopt-report=5 -qopt-report-phase=vec -qopt-report-phase=openmp
|
||||
# -guide -parallel
|
||||
# -guide-opts=string -guide-par[=n] -guide-vec[=n]
|
||||
# -auto-p32 -simd
|
||||
|
||||
# Reports: https://software.intel.com/en-us/articles/getting-the-most-out-of-your-intel-compiler-with-the-new-optimization-reports
|
||||
#CXXFLAGS += -qopt-report=5 -qopt-report-phase=vec,par
|
||||
#CXXFLAGS += -qopt-report=5 -qopt-report-phase=cg
|
||||
# Redirect report from *.optrpt to stderr
|
||||
# -qopt-report-file=stderr
|
||||
# Guided paralellization
|
||||
# -guide -parallel
|
||||
# -guide-opts=string -guide-par[=n] -guide-vec[=n]
|
||||
# -auto-p32 -simd
|
||||
|
||||
## run time checks
|
||||
# https://www.intel.com/content/www/us/en/develop/documentation/fortran-compiler-oneapi-dev-guide-and-reference/top/compiler-reference/compiler-options/offload-openmp-and-parallel-processing-options/par-runtime-control-qpar-runtime-control.html
|
||||
|
||||
|
||||
default: ${PROGRAM}
|
||||
|
||||
${PROGRAM}: ${OBJECTS}
|
||||
$(LINKER) $^ ${LINKFLAGS} -o $@
|
||||
|
||||
clean:
|
||||
rm -f ${PROGRAM} ${OBJECTS} *.optrpt
|
||||
|
||||
clean_all:: clean
|
||||
@rm -f *_ *~ *.bak *.log *.out *.tar
|
||||
|
||||
run: clean ${PROGRAM}
|
||||
./${PROGRAM}
|
||||
|
||||
# tar the current directory
|
||||
MY_DIR = `basename ${PWD}`
|
||||
tar: clean_all
|
||||
@echo "Tar the directory: " ${MY_DIR}
|
||||
@cd .. ;\
|
||||
tar cf ${MY_DIR}.tar ${MY_DIR} *default.mk ;\
|
||||
cd ${MY_DIR}
|
||||
# tar cf `basename ${PWD}`.tar *
|
||||
|
||||
doc:
|
||||
doxygen Doxyfile
|
||||
|
||||
#########################################################################
|
||||
|
||||
.cpp.o:
|
||||
$(CXX) -c $(CXXFLAGS) -o $@ $<
|
||||
|
||||
.c.o:
|
||||
$(CC) -c $(CFLAGS) -o $@ $<
|
||||
|
||||
.f.o:
|
||||
$(F77) -c $(FFLAGS) -o $@ $<
|
||||
|
||||
##################################################################################################
|
||||
# some tools
|
||||
# Cache behaviour (CXXFLAGS += -g tracks down to source lines)
|
||||
# https://software.intel.com/content/www/us/en/develop/documentation/vtune-help/top/analyze-performance/microarchitecture-analysis-group/memory-access-analysis.html
|
||||
|
||||
mem: inspector
|
||||
prof: vtune
|
||||
cache: inspector
|
||||
|
||||
gap_par_report:
|
||||
${CXX} -c -guide -parallel $(SOURCES) 2> gap.txt
|
||||
|
||||
# GUI for performance report
|
||||
amplifier: ${PROGRAM}
|
||||
echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope
|
||||
echo 0 | sudo tee /proc/sys/kernel/perf_event_paranoid
|
||||
amplxe-gui &
|
||||
|
||||
# GUI for Memory and Thread analyzer (race condition)
|
||||
inspector: ${PROGRAM}
|
||||
echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope
|
||||
# inspxe-gui &
|
||||
vtune-gui ./${PROGRAM} &
|
||||
|
||||
advisor:
|
||||
source /opt/intel/oneapi/advisor/2021.2.0/advixe-vars.sh
|
||||
# /opt/intel/oneapi/advisor/latest/bin64/advixe-gui &
|
||||
advisor --collect=survey ./${PROGRAM}
|
||||
# advisor --collect=roofline ./${PROGRAM}
|
||||
advisor --report=survey --project-dir=./ src:r=./ --format=csv --report-output=./out/survey.csv
|
||||
|
||||
vtune:
|
||||
echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope
|
||||
# https://software.intel.com/en-us/articles/intel-advisor-2017-update-1-what-s-new
|
||||
export ADVIXE_EXPERIMENTAL=roofline
|
||||
vtune -collect hotspots ./${PROGRAM}
|
||||
vtune -report hotspots -r r000hs > vtune.out
|
||||
# vtune-gui ./${PROGRAM} &
|
||||
|
||||
icc-info:
|
||||
icpc -# main.cpp
|
||||
|
||||
# MKL on AMD
|
||||
# https://www.computerbase.de/2019-11/mkl-workaround-erhoeht-leistung-auf-amd-ryzen/
|
||||
#
|
||||
# https://sites.google.com/a/uci.edu/mingru-yang/programming/mkl-has-bad-performance-on-an-amd-cpu
|
||||
# export MKL_DEBUG_CPU_TYPE=5
|
||||
# export MKL_NUM_THRAEDS=1
|
||||
# export MKL_DYNAMIC=false
|
||||
# on Intel compiler
|
||||
# http://publicclu2.blogspot.com/2013/05/intel-complier-suite-reference-card.html
|
||||
93
ex1/PGI_default.mk
Normal file
93
ex1/PGI_default.mk
Normal file
|
|
@ -0,0 +1,93 @@
|
|||
# Basic Defintions for using PGI-compiler suite sequentially
|
||||
# requires setting of COMPILER=PGI_
|
||||
# OPTIRUN = optirun
|
||||
|
||||
|
||||
CC = pgcc
|
||||
CXX = pgc++
|
||||
F77 = pgfortran
|
||||
LINKER = ${CXX}
|
||||
|
||||
# on mephisto:
|
||||
#CXXFLAGS += -I/share/apps/atlas/include
|
||||
#LINKFLAGS += -L/share/apps/atlas/lib
|
||||
#LINKFLAGS += -lcblas -latlas
|
||||
|
||||
#LINKFLAGS += -lblas
|
||||
# Der <cblas.h> Header muss mit extern "C" versehen werden, damit g++ alles findet.
|
||||
|
||||
WARNINGS = -Minform=warn
|
||||
# -Wall -pedantic -Wextra -Weffc++ -Woverloaded-virtual -W -Wfloat-equal -Wshadow -Wredundant-decls
|
||||
# -pedantic -Wunreachable-code -Wextra -Winline
|
||||
# -Wunreachable-code
|
||||
|
||||
#PGI_PROFILING = -Minfo=ccff,loop,vect,opt,intensity,mp,accel
|
||||
PGI_PROFILING = -Minfo=ccff,accel,ipa,loop,lre,mp,opt,par,unified,vect,intensity
|
||||
# -Minfo
|
||||
# -Mprof=time
|
||||
# -Mprof=lines
|
||||
# take care with option -Msafeptr
|
||||
CXXFLAGS += -O3 -std=c++17 ${WARNINGS}
|
||||
#CXXFLAGS += -O3 -std=c++11 -DNDEBUG ${PGI_PROFILING} ${WARNINGS}
|
||||
# -fastsse -fargument-noalias ${WARNINGS} -msse3 -vec-report=3
|
||||
|
||||
default: ${PROGRAM}
|
||||
|
||||
${PROGRAM}: ${OBJECTS}
|
||||
$(LINKER) $^ ${LINKFLAGS} -o $@
|
||||
|
||||
clean:
|
||||
@rm -f ${PROGRAM} ${OBJECTS}
|
||||
|
||||
clean_all:: clean
|
||||
@rm -f *_ *~ *.bak *.log *.out *.tar
|
||||
|
||||
run: clean ${PROGRAM}
|
||||
./${PROGRAM}
|
||||
|
||||
# tar the current directory
|
||||
MY_DIR = `basename ${PWD}`
|
||||
tar: clean_all
|
||||
@echo "Tar the directory: " ${MY_DIR}
|
||||
@cd .. ;\
|
||||
tar cf ${MY_DIR}.tar ${MY_DIR} *default.mk ;\
|
||||
cd ${MY_DIR}
|
||||
# tar cf `basename ${PWD}`.tar *
|
||||
|
||||
doc:
|
||||
doxygen Doxyfile
|
||||
|
||||
#########################################################################
|
||||
|
||||
.cpp.o:
|
||||
$(CXX) -c $(CXXFLAGS) -o $@ $<
|
||||
|
||||
.c.o:
|
||||
$(CC) -c $(CFLAGS) -o $@ $<
|
||||
|
||||
.f.o:
|
||||
$(F77) -c $(FFLAGS) -o $@ $<
|
||||
|
||||
##################################################################################################
|
||||
# # some tools
|
||||
# # Simple run time profiling of your code
|
||||
# # CXXFLAGS += -g -pg
|
||||
# # LINKFLAGS += -pg
|
||||
|
||||
|
||||
# Profiling options PGI, see: pgcollect -help
|
||||
# CPU_PROF = -allcache
|
||||
CPU_PROF = -time
|
||||
# GPU_PROF = -cuda=gmem,branch,cc13 -cudainit
|
||||
#GPU_PROF = -cuda=branch:cc20
|
||||
#
|
||||
PROF_FILE = pgprof.out
|
||||
|
||||
cache: prof
|
||||
|
||||
prof: ${PROGRAM}
|
||||
${OPTIRUN} ${BINDIR}pgcollect $(CPU_PROF) ./$^
|
||||
${OPTIRUN} ${BINDIR}pgprof -exe ./$^ $(PROF_FILE) &
|
||||
|
||||
info:
|
||||
pgaccelinfo -v
|
||||
28
ex1/ex1A_mean_values/.vscode/tasks.json
vendored
Normal file
28
ex1/ex1A_mean_values/.vscode/tasks.json
vendored
Normal file
|
|
@ -0,0 +1,28 @@
|
|||
{
|
||||
"tasks": [
|
||||
{
|
||||
"type": "cppbuild",
|
||||
"label": "C/C++: gcc build active file",
|
||||
"command": "/usr/bin/gcc",
|
||||
"args": [
|
||||
"-fdiagnostics-color=always",
|
||||
"-g",
|
||||
"${file}",
|
||||
"-o",
|
||||
"${fileDirname}/${fileBasenameNoExtension}"
|
||||
],
|
||||
"options": {
|
||||
"cwd": "${fileDirname}"
|
||||
},
|
||||
"problemMatcher": [
|
||||
"$gcc"
|
||||
],
|
||||
"group": {
|
||||
"kind": "build",
|
||||
"isDefault": true
|
||||
},
|
||||
"detail": "Task generated by Debugger."
|
||||
}
|
||||
],
|
||||
"version": "2.0.0"
|
||||
}
|
||||
30
ex1/ex1A_mean_values/Makefile
Normal file
30
ex1/ex1A_mean_values/Makefile
Normal file
|
|
@ -0,0 +1,30 @@
|
|||
#
|
||||
# use GNU-Compiler tools
|
||||
COMPILER=GCC_
|
||||
# alternatively from the shell
|
||||
# export COMPILER=GCC_
|
||||
# or, alternatively from the shell
|
||||
# make COMPILER=GCC_
|
||||
|
||||
# use Intel compilers
|
||||
#COMPILER=ICC_
|
||||
|
||||
# use PGI compilers
|
||||
# COMPILER=PGI_
|
||||
|
||||
|
||||
SOURCES = main.cpp ../ex1A_mean_values/means.cpp
|
||||
OBJECTS = $(SOURCES:.cpp=.o)
|
||||
|
||||
PROGRAM = main.${COMPILER}
|
||||
|
||||
# uncomment the next to lines for debugging and detailed performance analysis
|
||||
CXXFLAGS += -g
|
||||
LINKFLAGS += -g
|
||||
# do not use -pg with PGI compilers
|
||||
|
||||
ifndef COMPILER
|
||||
COMPILER=GCC_
|
||||
endif
|
||||
|
||||
include ../${COMPILER}default.mk
|
||||
BIN
ex1/ex1A_mean_values/main.GCC_
Executable file
BIN
ex1/ex1A_mean_values/main.GCC_
Executable file
Binary file not shown.
35
ex1/ex1A_mean_values/main.cpp
Normal file
35
ex1/ex1A_mean_values/main.cpp
Normal file
|
|
@ -0,0 +1,35 @@
|
|||
#include "means.h"
|
||||
#include <vector>
|
||||
#include <iostream>
|
||||
using namespace std;
|
||||
|
||||
int main(int argc, char **argv)
|
||||
{
|
||||
double arithmetic_mean, geometric_mean, harmonic_mean;
|
||||
|
||||
// Fixed version
|
||||
calculate_means(1, 4, 16, arithmetic_mean, geometric_mean, harmonic_mean);
|
||||
cout << arithmetic_mean << ", " << geometric_mean << ", " << harmonic_mean << endl;
|
||||
|
||||
calculate_means(2, 3, 5, arithmetic_mean, geometric_mean, harmonic_mean);
|
||||
cout << arithmetic_mean << ", " << geometric_mean << ", " << harmonic_mean << endl;
|
||||
|
||||
calculate_means(1000, 4000, 16000, arithmetic_mean, geometric_mean, harmonic_mean);
|
||||
cout << arithmetic_mean << ", " << geometric_mean << ", " << harmonic_mean << endl;
|
||||
cout << "--------------------------------" << endl;
|
||||
|
||||
|
||||
|
||||
// Scalable version
|
||||
calculate_means(vector<int> {1, 4, 16}, arithmetic_mean, geometric_mean, harmonic_mean);
|
||||
cout << arithmetic_mean << ", " << geometric_mean << ", " << harmonic_mean << endl;
|
||||
|
||||
calculate_means(vector<int> {2, 3, 5}, arithmetic_mean, geometric_mean, harmonic_mean);
|
||||
cout << arithmetic_mean << ", " << geometric_mean << ", " << harmonic_mean << endl;
|
||||
|
||||
calculate_means(vector<int> {1000, 4000, 16000}, arithmetic_mean, geometric_mean, harmonic_mean);
|
||||
cout << arithmetic_mean << ", " << geometric_mean << ", " << harmonic_mean << endl;
|
||||
|
||||
|
||||
return 0;
|
||||
}
|
||||
BIN
ex1/ex1A_mean_values/main.o
Normal file
BIN
ex1/ex1A_mean_values/main.o
Normal file
Binary file not shown.
30
ex1/ex1A_mean_values/means.cpp
Normal file
30
ex1/ex1A_mean_values/means.cpp
Normal file
|
|
@ -0,0 +1,30 @@
|
|||
#include "../ex1A_mean_values/means.h"
|
||||
#include <cmath>
|
||||
#include <vector>
|
||||
|
||||
void calculate_means(int a, int b, int c, double &am, double &gm, double &hm)
|
||||
{
|
||||
am = (a + b + c)/3.0;
|
||||
gm = exp((log(a)+log(b)+log(c))/3);
|
||||
hm = 3.0/(1.0/a + 1.0/b + 1.0/c);
|
||||
}
|
||||
|
||||
void calculate_means(std::vector<int> numbers, double &am, double &gm, double &hm)
|
||||
{
|
||||
int n = numbers.size();
|
||||
|
||||
am = 0.;
|
||||
gm = 0.;
|
||||
hm = 0.;
|
||||
|
||||
for (int i = 0; i < n; ++i)
|
||||
{
|
||||
am += numbers[i];
|
||||
gm += log(numbers[i]);
|
||||
hm += 1.0/numbers[i];
|
||||
}
|
||||
|
||||
am /= n;
|
||||
gm = exp(gm/n);
|
||||
hm = n/hm;
|
||||
}
|
||||
23
ex1/ex1A_mean_values/means.h
Normal file
23
ex1/ex1A_mean_values/means.h
Normal file
|
|
@ -0,0 +1,23 @@
|
|||
#pragma once
|
||||
#include <vector>
|
||||
|
||||
/**
|
||||
This function calculates arithmetic mean, geometric mean and harmonic mean of three integers.
|
||||
@param[in] a first integer
|
||||
@param[in] b second integer
|
||||
@param[in] c third integer
|
||||
@param[out] am arithmetic mean
|
||||
@param[out] gm geometric mean
|
||||
@param[out] hm harmonic mean
|
||||
*/
|
||||
void calculate_means(int a, int b, int c, double &am, double &gm, double &hm);
|
||||
|
||||
/**
|
||||
This function calculates arithmetic mean, geometric mean and harmonic mean of an integer vector.
|
||||
@param[in] numbers vector containing integers
|
||||
@param[out] am arithmetic mean
|
||||
@param[out] gm geometric mean
|
||||
@param[out] hm harmonic mean
|
||||
*/
|
||||
void calculate_means(std::vector<int> numbers, double &am, double &gm, double &hm);
|
||||
|
||||
BIN
ex1/ex1A_mean_values/means.o
Normal file
BIN
ex1/ex1A_mean_values/means.o
Normal file
Binary file not shown.
30
ex1/ex1B_data-IO_and_vectors/Makefile
Normal file
30
ex1/ex1B_data-IO_and_vectors/Makefile
Normal file
|
|
@ -0,0 +1,30 @@
|
|||
#
|
||||
# use GNU-Compiler tools
|
||||
COMPILER=GCC_
|
||||
# alternatively from the shell
|
||||
# export COMPILER=GCC_
|
||||
# or, alternatively from the shell
|
||||
# make COMPILER=GCC_
|
||||
|
||||
# use Intel compilers
|
||||
#COMPILER=ICC_
|
||||
|
||||
# use PGI compilers
|
||||
# COMPILER=PGI_
|
||||
|
||||
|
||||
SOURCES = main.cpp ../ex1A_mean_values/means.cpp
|
||||
OBJECTS = $(SOURCES:.cpp=.o)
|
||||
|
||||
PROGRAM = main.${COMPILER}
|
||||
|
||||
# uncomment the next to lines for debugging and detailed performance analysis
|
||||
CXXFLAGS += -g
|
||||
LINKFLAGS += -g
|
||||
# do not use -pg with PGI compilers
|
||||
|
||||
ifndef COMPILER
|
||||
COMPILER=GCC_
|
||||
endif
|
||||
|
||||
include ../${COMPILER}default.mk
|
||||
501
ex1/ex1B_data-IO_and_vectors/data_1.txt
Normal file
501
ex1/ex1B_data-IO_and_vectors/data_1.txt
Normal file
|
|
@ -0,0 +1,501 @@
|
|||
141
|
||||
261
|
||||
87
|
||||
430
|
||||
258
|
||||
298
|
||||
425
|
||||
120
|
||||
496
|
||||
707
|
||||
244
|
||||
786
|
||||
75
|
||||
394
|
||||
4
|
||||
221
|
||||
2
|
||||
190
|
||||
143
|
||||
269
|
||||
175
|
||||
139
|
||||
599
|
||||
902
|
||||
940
|
||||
222
|
||||
483
|
||||
377
|
||||
524
|
||||
265
|
||||
69
|
||||
437
|
||||
174
|
||||
27
|
||||
955
|
||||
431
|
||||
962
|
||||
763
|
||||
8
|
||||
681
|
||||
706
|
||||
646
|
||||
553
|
||||
219
|
||||
773
|
||||
229
|
||||
371
|
||||
891
|
||||
857
|
||||
403
|
||||
319
|
||||
609
|
||||
911
|
||||
910
|
||||
592
|
||||
333
|
||||
854
|
||||
443
|
||||
905
|
||||
34
|
||||
533
|
||||
717
|
||||
180
|
||||
337
|
||||
188
|
||||
322
|
||||
404
|
||||
549
|
||||
49
|
||||
553
|
||||
275
|
||||
242
|
||||
244
|
||||
155
|
||||
957
|
||||
936
|
||||
819
|
||||
729
|
||||
176
|
||||
361
|
||||
189
|
||||
2
|
||||
317
|
||||
700
|
||||
626
|
||||
544
|
||||
440
|
||||
288
|
||||
502
|
||||
762
|
||||
763
|
||||
577
|
||||
748
|
||||
646
|
||||
124
|
||||
505
|
||||
348
|
||||
93
|
||||
148
|
||||
199
|
||||
673
|
||||
432
|
||||
695
|
||||
257
|
||||
10
|
||||
533
|
||||
280
|
||||
947
|
||||
907
|
||||
393
|
||||
25
|
||||
672
|
||||
838
|
||||
972
|
||||
57
|
||||
451
|
||||
583
|
||||
687
|
||||
720
|
||||
651
|
||||
727
|
||||
374
|
||||
582
|
||||
117
|
||||
58
|
||||
980
|
||||
285
|
||||
595
|
||||
963
|
||||
186
|
||||
194
|
||||
342
|
||||
933
|
||||
391
|
||||
274
|
||||
152
|
||||
398
|
||||
375
|
||||
132
|
||||
436
|
||||
92
|
||||
615
|
||||
11
|
||||
574
|
||||
790
|
||||
236
|
||||
449
|
||||
570
|
||||
62
|
||||
497
|
||||
643
|
||||
222
|
||||
838
|
||||
972
|
||||
847
|
||||
506
|
||||
279
|
||||
747
|
||||
237
|
||||
958
|
||||
621
|
||||
601
|
||||
173
|
||||
91
|
||||
256
|
||||
859
|
||||
912
|
||||
700
|
||||
726
|
||||
230
|
||||
577
|
||||
811
|
||||
404
|
||||
989
|
||||
90
|
||||
321
|
||||
512
|
||||
61
|
||||
726
|
||||
557
|
||||
530
|
||||
830
|
||||
859
|
||||
790
|
||||
318
|
||||
453
|
||||
753
|
||||
110
|
||||
110
|
||||
270
|
||||
525
|
||||
973
|
||||
711
|
||||
312
|
||||
292
|
||||
851
|
||||
912
|
||||
640
|
||||
256
|
||||
89
|
||||
839
|
||||
585
|
||||
949
|
||||
62
|
||||
585
|
||||
286
|
||||
828
|
||||
191
|
||||
443
|
||||
394
|
||||
827
|
||||
677
|
||||
208
|
||||
319
|
||||
134
|
||||
672
|
||||
571
|
||||
170
|
||||
148
|
||||
477
|
||||
909
|
||||
553
|
||||
33
|
||||
54
|
||||
806
|
||||
452
|
||||
383
|
||||
790
|
||||
365
|
||||
533
|
||||
712
|
||||
872
|
||||
329
|
||||
651
|
||||
975
|
||||
76
|
||||
588
|
||||
414
|
||||
310
|
||||
264
|
||||
759
|
||||
996
|
||||
187
|
||||
782
|
||||
196
|
||||
993
|
||||
803
|
||||
425
|
||||
729
|
||||
499
|
||||
809
|
||||
357
|
||||
74
|
||||
591
|
||||
911
|
||||
194
|
||||
433
|
||||
750
|
||||
40
|
||||
947
|
||||
764
|
||||
559
|
||||
184
|
||||
498
|
||||
518
|
||||
995
|
||||
855
|
||||
963
|
||||
679
|
||||
404
|
||||
935
|
||||
480
|
||||
232
|
||||
397
|
||||
706
|
||||
559
|
||||
757
|
||||
996
|
||||
963
|
||||
536
|
||||
964
|
||||
116
|
||||
52
|
||||
305
|
||||
581
|
||||
531
|
||||
902
|
||||
541
|
||||
432
|
||||
543
|
||||
713
|
||||
17
|
||||
801
|
||||
143
|
||||
479
|
||||
257
|
||||
370
|
||||
662
|
||||
170
|
||||
279
|
||||
199
|
||||
196
|
||||
327
|
||||
881
|
||||
472
|
||||
404
|
||||
180
|
||||
969
|
||||
408
|
||||
845
|
||||
616
|
||||
377
|
||||
878
|
||||
785
|
||||
465
|
||||
814
|
||||
899
|
||||
430
|
||||
335
|
||||
597
|
||||
902
|
||||
703
|
||||
378
|
||||
735
|
||||
955
|
||||
543
|
||||
541
|
||||
312
|
||||
72
|
||||
182
|
||||
93
|
||||
464
|
||||
10
|
||||
916
|
||||
643
|
||||
2
|
||||
31
|
||||
209
|
||||
455
|
||||
128
|
||||
9
|
||||
728
|
||||
355
|
||||
781
|
||||
437
|
||||
437
|
||||
50
|
||||
50
|
||||
92
|
||||
595
|
||||
242
|
||||
842
|
||||
858
|
||||
964
|
||||
489
|
||||
221
|
||||
227
|
||||
537
|
||||
763
|
||||
348
|
||||
462
|
||||
640
|
||||
918
|
||||
162
|
||||
716
|
||||
578
|
||||
434
|
||||
885
|
||||
394
|
||||
179
|
||||
634
|
||||
625
|
||||
328
|
||||
803
|
||||
1000
|
||||
981
|
||||
128
|
||||
233
|
||||
24
|
||||
608
|
||||
111
|
||||
408
|
||||
885
|
||||
549
|
||||
370
|
||||
209
|
||||
441
|
||||
957
|
||||
125
|
||||
471
|
||||
857
|
||||
44
|
||||
692
|
||||
979
|
||||
284
|
||||
134
|
||||
686
|
||||
910
|
||||
611
|
||||
900
|
||||
194
|
||||
755
|
||||
347
|
||||
419
|
||||
156
|
||||
820
|
||||
625
|
||||
739
|
||||
806
|
||||
68
|
||||
951
|
||||
498
|
||||
756
|
||||
743
|
||||
832
|
||||
157
|
||||
458
|
||||
619
|
||||
933
|
||||
836
|
||||
896
|
||||
583
|
||||
583
|
||||
855
|
||||
35
|
||||
886
|
||||
408
|
||||
37
|
||||
747
|
||||
155
|
||||
144
|
||||
606
|
||||
255
|
||||
325
|
||||
402
|
||||
407
|
||||
387
|
||||
610
|
||||
167
|
||||
189
|
||||
95
|
||||
324
|
||||
770
|
||||
235
|
||||
741
|
||||
693
|
||||
825
|
||||
828
|
||||
294
|
||||
310
|
||||
524
|
||||
326
|
||||
832
|
||||
811
|
||||
557
|
||||
263
|
||||
681
|
||||
234
|
||||
457
|
||||
385
|
||||
539
|
||||
992
|
||||
756
|
||||
981
|
||||
235
|
||||
529
|
||||
52
|
||||
757
|
||||
602
|
||||
858
|
||||
989
|
||||
930
|
||||
410
|
||||
1
|
||||
541
|
||||
208
|
||||
220
|
||||
326
|
||||
96
|
||||
748
|
||||
749
|
||||
544
|
||||
339
|
||||
833
|
||||
553
|
||||
958
|
||||
893
|
||||
357
|
||||
547
|
||||
347
|
||||
623
|
||||
797
|
||||
746
|
||||
126
|
||||
823
|
||||
26
|
||||
415
|
||||
732
|
||||
782
|
||||
368
|
||||
|
||||
BIN
ex1/ex1B_data-IO_and_vectors/main.GCC_
Executable file
BIN
ex1/ex1B_data-IO_and_vectors/main.GCC_
Executable file
Binary file not shown.
53
ex1/ex1B_data-IO_and_vectors/main.cpp
Normal file
53
ex1/ex1B_data-IO_and_vectors/main.cpp
Normal file
|
|
@ -0,0 +1,53 @@
|
|||
#include "../ex1A_mean_values/means.h"
|
||||
#include <iostream>
|
||||
#include <fstream>
|
||||
#include <cmath>
|
||||
#include <vector>
|
||||
#include <algorithm>
|
||||
using namespace std;
|
||||
|
||||
|
||||
int main(int argc, char **argv)
|
||||
{
|
||||
// read vector from file
|
||||
vector<int> data_vector = {};
|
||||
|
||||
ifstream input_stream("data_1.txt");
|
||||
|
||||
int line;
|
||||
while(input_stream >> line)
|
||||
{
|
||||
data_vector.push_back(line);
|
||||
}
|
||||
data_vector.shrink_to_fit();
|
||||
|
||||
|
||||
// calculate minimum and maximum
|
||||
vector<int>::iterator min_it = min_element(data_vector.begin(), data_vector.end());
|
||||
vector<int>::iterator max_it = max_element(data_vector.begin(), data_vector.end());
|
||||
|
||||
// calculate arithmetic mean, geometric mean and harmonic mean
|
||||
double am, gm, hm;
|
||||
calculate_means(data_vector, am, gm, hm);
|
||||
|
||||
|
||||
// calculate standard deviation
|
||||
double sd = 0.;
|
||||
int n = data_vector.size();
|
||||
for (int i = 0; i < n; ++i)
|
||||
{
|
||||
sd += pow(data_vector[i] - am, 2);
|
||||
}
|
||||
sd = sqrt(sd/n);
|
||||
|
||||
|
||||
// print results
|
||||
cout << "minimum: " << *min_it << endl;
|
||||
cout << "maximum: " << *max_it << endl;
|
||||
cout << "arithmetic mean: " << am << endl;
|
||||
cout << "geometric mean: " << gm << endl;
|
||||
cout << "harmonic mean: " << hm << endl;
|
||||
cout << "standard deviation: " << sd << endl;
|
||||
|
||||
return 0;
|
||||
}
|
||||
BIN
ex1/ex1B_data-IO_and_vectors/main.o
Normal file
BIN
ex1/ex1B_data-IO_and_vectors/main.o
Normal file
Binary file not shown.
31
ex1/ex1C_summation_of_specified_numbers/main.cpp
Normal file
31
ex1/ex1C_summation_of_specified_numbers/main.cpp
Normal file
|
|
@ -0,0 +1,31 @@
|
|||
#include "special_sum.h"
|
||||
#include "../utils/timing.h"
|
||||
#include <iostream>
|
||||
#include <chrono>
|
||||
#include <stack>
|
||||
using namespace std;
|
||||
|
||||
int main(int argc, char **argv)
|
||||
{
|
||||
// check results and compare speeds
|
||||
for(size_t n : {15, 1001, 1432987})
|
||||
{
|
||||
cout << "n = " << n << endl;
|
||||
size_t sum_1, sum_2;
|
||||
|
||||
tic();
|
||||
for(size_t i = 0; i < 1000; ++i)
|
||||
sum_1 = special_sum_loop(n);
|
||||
double time_1 = toc();
|
||||
|
||||
tic();
|
||||
for(size_t i = 0; i < 1000; ++i)
|
||||
sum_2 = special_sum_noloop(n);
|
||||
double time_2 = toc();
|
||||
|
||||
cout << "loop: " << sum_1 << "\t\tDuration: " << time_1 << endl;
|
||||
cout << "no loop: " << sum_2 << "\t\tDuration: " << time_2 << endl << "---------------------------------------------------" << endl;
|
||||
}
|
||||
|
||||
return 0;
|
||||
}
|
||||
28
ex1/ex1C_summation_of_specified_numbers/special_sum.cpp
Normal file
28
ex1/ex1C_summation_of_specified_numbers/special_sum.cpp
Normal file
|
|
@ -0,0 +1,28 @@
|
|||
#include "special_sum.h"
|
||||
|
||||
size_t gauss_sum(size_t n)
|
||||
{
|
||||
return (n*(n+1))/2;
|
||||
}
|
||||
|
||||
size_t special_sum_loop(size_t n)
|
||||
{
|
||||
size_t sum = 0;
|
||||
for (size_t i = 1; i < n+1; ++i)
|
||||
{
|
||||
if (i % 3 == 0 || i % 5 == 0)
|
||||
{
|
||||
sum += i;
|
||||
}
|
||||
}
|
||||
return sum;
|
||||
}
|
||||
|
||||
size_t special_sum_noloop(size_t n)
|
||||
{
|
||||
size_t factor_3 = gauss_sum(n/3); // dividing int by int automatically gets rounded off
|
||||
size_t factor_5 = gauss_sum(n/5);
|
||||
size_t factor_15 = gauss_sum(n/15);
|
||||
|
||||
return factor_3*3 + factor_5*5 - factor_15*15;
|
||||
}
|
||||
18
ex1/ex1C_summation_of_specified_numbers/special_sum.h
Normal file
18
ex1/ex1C_summation_of_specified_numbers/special_sum.h
Normal file
|
|
@ -0,0 +1,18 @@
|
|||
#include <cstddef>
|
||||
|
||||
/**
|
||||
This function returns the sum of all positive integers less or equal n which are a multiples of 3 or of 5, WITH using a loop.
|
||||
@param[in] n
|
||||
@param[out] M
|
||||
*/
|
||||
size_t special_sum_loop(size_t n);
|
||||
|
||||
|
||||
/**
|
||||
This function returns the sum of all positive integers less or equal n which are a multiples of 3 or of 5, WITHOUT using a loop.
|
||||
Example: For n=15, we have 60 = 3+5+6+9+10+12+15 = (1+2+3+4+5)*3 + (1+2+3)*5 - 1*15
|
||||
Formula: M = (\sum_{i=1}^{k_3} i)*3 + (\sum_{i=1}^{k_5} i)*5 - (\sum_{i=1}^{k_15} i)*15
|
||||
@param[in] n
|
||||
@param[out] M
|
||||
*/
|
||||
size_t special_sum_noloop(size_t n);
|
||||
BIN
ex1/ex1D_kahan_summation/main
Executable file
BIN
ex1/ex1D_kahan_summation/main
Executable file
Binary file not shown.
33
ex1/ex1D_kahan_summation/main.cpp
Normal file
33
ex1/ex1D_kahan_summation/main.cpp
Normal file
|
|
@ -0,0 +1,33 @@
|
|||
#include "mylib.h"
|
||||
#include <cmath>
|
||||
#include <iostream>
|
||||
using namespace std;
|
||||
|
||||
int main(int argc, char **argv)
|
||||
{
|
||||
for(size_t i = 1; i < 8; ++i)
|
||||
{
|
||||
size_t n = pow(10,i);
|
||||
vector<double> x(n);
|
||||
for (size_t k = 0; k < n; ++k)
|
||||
x[k] = 1.0/(k + 1);
|
||||
|
||||
|
||||
// compute scalar products
|
||||
double sum_1 = scalar(x, x);
|
||||
double sum_2 = Kahan_skalar(x, x);
|
||||
|
||||
// compute error
|
||||
double err_1 = abs(sum_1 - pow(M_PI,2)/6);
|
||||
double err_2 = abs(sum_2 - pow(M_PI,2)/6);
|
||||
|
||||
cout << "n = " << n << endl;
|
||||
cout << "Normal scalar product: " << sum_1 << "\terror: " << err_1 << endl;
|
||||
cout << "Kahan scalar product: " << sum_2 << "\terror: " << err_2 << endl;
|
||||
cout << endl;
|
||||
}
|
||||
|
||||
|
||||
|
||||
return 0;
|
||||
}
|
||||
34
ex1/ex1D_kahan_summation/mylib.cpp
Normal file
34
ex1/ex1D_kahan_summation/mylib.cpp
Normal file
|
|
@ -0,0 +1,34 @@
|
|||
#include "mylib.h"
|
||||
#include <cassert> // assert()
|
||||
#include <cmath>
|
||||
#include <vector>
|
||||
using namespace std;
|
||||
|
||||
double scalar(vector<double> const &x, vector<double> const &y)
|
||||
{
|
||||
assert(x.size() == y.size());
|
||||
size_t const N = x.size();
|
||||
double sum = 0.0;
|
||||
for (size_t i = 0; i < N; ++i)
|
||||
{
|
||||
sum += x[i] * y[i];
|
||||
}
|
||||
return sum;
|
||||
}
|
||||
|
||||
|
||||
|
||||
double Kahan_skalar(vector<double> const &x, vector<double> const &y)
|
||||
{
|
||||
double sum = 0;
|
||||
double c = 0;
|
||||
size_t n = x.size();
|
||||
for (size_t i = 0; i < n; ++i)
|
||||
{
|
||||
double z = x[i]*y[i] - c; // c is the part that got lost in the last iteration
|
||||
double t = sum + z; // when adding sum + z, the lower digits are lost if sum is large
|
||||
c = (t - sum) - z; // now we recover the lower digits to add in the next iteration
|
||||
sum = t;
|
||||
}
|
||||
return sum;
|
||||
}
|
||||
30
ex1/ex1D_kahan_summation/mylib.h
Normal file
30
ex1/ex1D_kahan_summation/mylib.h
Normal file
|
|
@ -0,0 +1,30 @@
|
|||
#ifndef FILE_MYLIB
|
||||
#define FILE_MYLIB
|
||||
#include <vector>
|
||||
|
||||
/** Inner product
|
||||
@param[in] x vector
|
||||
@param[in] y vector
|
||||
@return resulting Euclidian inner product <x,y>
|
||||
*/
|
||||
double scalar(std::vector<double> const &x, std::vector<double> const &y);
|
||||
|
||||
/** Inner product using BLAS routines
|
||||
@param[in] x vector
|
||||
@param[in] y vector
|
||||
@return resulting Euclidian inner product <x,y>
|
||||
*/
|
||||
double scalar_cblas(std::vector<double> const &x, std::vector<double> const &y);
|
||||
float scalar_cblas(std::vector<float> const &x, std::vector<float> const &y);
|
||||
|
||||
|
||||
/** L_2 Norm of a vector
|
||||
@param[in] x vector
|
||||
@return resulting Euclidian norm <x,y>
|
||||
*/
|
||||
double norm(std::vector<double> const &x);
|
||||
|
||||
double Kahan_skalar(std::vector<double> const &x, std::vector<double> const &y);
|
||||
|
||||
|
||||
#endif
|
||||
BIN
ex1/ex1E_vector_vs_list/main
Executable file
BIN
ex1/ex1E_vector_vs_list/main
Executable file
Binary file not shown.
63
ex1/ex1E_vector_vs_list/main.cpp
Normal file
63
ex1/ex1E_vector_vs_list/main.cpp
Normal file
|
|
@ -0,0 +1,63 @@
|
|||
#include "../utils/timing.h"
|
||||
#include <iostream>
|
||||
#include <random>
|
||||
#include <chrono>
|
||||
#include <vector>
|
||||
#include <list>
|
||||
#include <algorithm>
|
||||
using namespace std;
|
||||
|
||||
|
||||
size_t random_integer(int lower_bound, int upper_bound)
|
||||
{
|
||||
unsigned seed = chrono::system_clock::now().time_since_epoch().count();
|
||||
|
||||
minstd_rand0 generator (seed);
|
||||
|
||||
return lower_bound + generator() % (upper_bound - lower_bound + 1);
|
||||
}
|
||||
|
||||
int main(int argc, char **argv)
|
||||
{
|
||||
// start with generating a sorted vector/list
|
||||
size_t n = 10000;
|
||||
vector<int> x_vec(n);
|
||||
list<int> x_list(n);
|
||||
for(size_t k = 0; k < n; ++k)
|
||||
{
|
||||
x_vec[k] = k + 1;
|
||||
x_list.push_back(k + 1);
|
||||
}
|
||||
|
||||
|
||||
// insert new random entries such that the container stays sorted
|
||||
tic();
|
||||
for(size_t i = 0; i < n; ++i)
|
||||
{
|
||||
size_t new_entry = random_integer(1,n);
|
||||
auto it = lower_bound(x_vec.begin(), x_vec.end(), new_entry);
|
||||
x_vec.insert(it, new_entry);
|
||||
}
|
||||
double time_1 = toc();
|
||||
|
||||
tic();
|
||||
for(size_t i = 0; i < n; ++i)
|
||||
{
|
||||
size_t new_entry = random_integer(1,n);
|
||||
auto it = lower_bound(x_list.begin(), x_list.end(), new_entry);
|
||||
x_list.insert(it, new_entry);
|
||||
}
|
||||
double time_2 = toc();
|
||||
|
||||
|
||||
// check results
|
||||
cout << "New vector is sorted: " << std::boolalpha << is_sorted(x_vec.cbegin(), x_vec.cend()) << "\tsize: " << x_vec.size() << "\tduration: " << time_1 << endl;
|
||||
cout << "New list is sorted: " << std::boolalpha << is_sorted(x_list.cbegin(), x_list.cend()) << "\tsize: " << x_list.size() << "\tduration: " << time_2 << endl;
|
||||
|
||||
// Vector stores 3 pointers
|
||||
// List stores two pointers for every element: one to the previous, one to the next element
|
||||
|
||||
|
||||
|
||||
return 0;
|
||||
}
|
||||
BIN
ex1/ex1F_goldbachs_conjecture/goldbach
Executable file
BIN
ex1/ex1F_goldbachs_conjecture/goldbach
Executable file
Binary file not shown.
42
ex1/ex1F_goldbachs_conjecture/goldbach.cpp
Normal file
42
ex1/ex1F_goldbachs_conjecture/goldbach.cpp
Normal file
|
|
@ -0,0 +1,42 @@
|
|||
#include "goldbach.h"
|
||||
|
||||
size_t single_goldbach(size_t k)
|
||||
{
|
||||
const std::vector<size_t> relevant_primes = get_primes(k);
|
||||
size_t m = relevant_primes.size();
|
||||
|
||||
size_t counter = 0;
|
||||
|
||||
for(size_t i = 0; i < m; ++i)
|
||||
{
|
||||
for(size_t j = i; j < m; ++j)
|
||||
{
|
||||
if(relevant_primes[i] + relevant_primes[j] == k)
|
||||
++counter;
|
||||
}
|
||||
}
|
||||
|
||||
return counter;
|
||||
}
|
||||
|
||||
|
||||
std::vector<size_t> count_goldbach(size_t n)
|
||||
{
|
||||
const std::vector<size_t> relevant_primes = get_primes(n);
|
||||
size_t m = relevant_primes.size();
|
||||
|
||||
std::vector<size_t> counter_vector(n + 1, 0);
|
||||
|
||||
|
||||
for(size_t i = 0; i < m; ++i)
|
||||
{
|
||||
for(size_t j = i; j < m; ++j)
|
||||
{
|
||||
size_t sum = relevant_primes[i] + relevant_primes[j];
|
||||
if(sum <= n)
|
||||
++counter_vector[relevant_primes[i] + relevant_primes[j]];
|
||||
}
|
||||
}
|
||||
|
||||
return counter_vector;
|
||||
}
|
||||
21
ex1/ex1F_goldbachs_conjecture/goldbach.h
Normal file
21
ex1/ex1F_goldbachs_conjecture/goldbach.h
Normal file
|
|
@ -0,0 +1,21 @@
|
|||
#pragma once
|
||||
#include "mayer_primes.h"
|
||||
#include <iostream>
|
||||
#include <vector>
|
||||
#include <iterator>
|
||||
#include <cassert>
|
||||
|
||||
|
||||
/**
|
||||
This function returns the number of possible decompositions of an integer into a sum of two prime numbers.
|
||||
@param[in] k first integer
|
||||
@param[out] count number of decompositions
|
||||
*/
|
||||
size_t single_goldbach(size_t k);
|
||||
|
||||
/**
|
||||
This function returns the number of possible decompositions into a sum of two prime numbers of all even integers in the interval [4,n].
|
||||
@param[in] n upper integer bound
|
||||
@param[out] count_vector vector containing the number of decompositions for a natural number the corresponding index
|
||||
*/
|
||||
std::vector<size_t> count_goldbach(size_t n);
|
||||
BIN
ex1/ex1F_goldbachs_conjecture/main
Executable file
BIN
ex1/ex1F_goldbachs_conjecture/main
Executable file
Binary file not shown.
37
ex1/ex1F_goldbachs_conjecture/main.cpp
Normal file
37
ex1/ex1F_goldbachs_conjecture/main.cpp
Normal file
|
|
@ -0,0 +1,37 @@
|
|||
#include "../utils/timing.h"
|
||||
#include "goldbach.h"
|
||||
#include <iostream>
|
||||
#include <algorithm>
|
||||
using namespace std;
|
||||
|
||||
|
||||
int main(int argc, char **argv)
|
||||
{
|
||||
cout << "Check: 694 has "<< single_goldbach(694) << " decompositions." << endl << "----------------------------------------" << endl;
|
||||
|
||||
|
||||
for(size_t n : {10000, 100000, 400000, 1000000, 2000000})
|
||||
{
|
||||
tic();
|
||||
|
||||
auto goldbach_vector = count_goldbach(n);
|
||||
|
||||
auto max_it = max_element(goldbach_vector.begin(), goldbach_vector.end());
|
||||
size_t max_number = distance(goldbach_vector.begin(), max_it);
|
||||
|
||||
double time = toc();
|
||||
|
||||
cout << "The number " << max_number << " has " << *max_it << " decompositions. Duration: " << time << endl;
|
||||
}
|
||||
|
||||
/*
|
||||
The number 9240 has 329 decompositions. Duration: 0.00572876
|
||||
The number 99330 has 2168 decompositions. Duration: 0.3342
|
||||
The number 390390 has 7094 decompositions. Duration: 4.23734
|
||||
The number 990990 has 15594 decompositions. Duration: 29.5817
|
||||
The number 1981980 has 27988 decompositions. Duration: 135.985
|
||||
*/
|
||||
|
||||
|
||||
return 0;
|
||||
}
|
||||
73
ex1/ex1F_goldbachs_conjecture/mayer_primes.h
Normal file
73
ex1/ex1F_goldbachs_conjecture/mayer_primes.h
Normal file
|
|
@ -0,0 +1,73 @@
|
|||
#pragma once
|
||||
|
||||
#include <cstring> //memset
|
||||
#include <vector>
|
||||
//using namespace std;
|
||||
|
||||
/** \brief Determines all prime numbers in interval [2, @p max].
|
||||
*
|
||||
* The sieve of Eratosthenes is used.
|
||||
*
|
||||
* The implementation originates from <a href="http://code.activestate.com/recipes/576559-fast-prime-generator/">Florian Mayer</a>.
|
||||
*
|
||||
* \param[in] max end of interval for the prime number search.
|
||||
* \return vector of prime numbers @f$2,3,5, ..., p<=max @f$.
|
||||
*
|
||||
* \copyright
|
||||
* Copyright (c) 2008 Florian Mayer (adapted by Gundolf Haase 2018)
|
||||
*
|
||||
* Permission is hereby granted, free of charge, to any person obtaining a copy
|
||||
* of this software and associated documentation files (the "Software"), to deal
|
||||
* in the Software without restriction, including without limitation the rights
|
||||
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
||||
* copies of the Software, and to permit persons to whom the Software is
|
||||
* furnished to do so, subject to the following conditions:
|
||||
*
|
||||
* The above copyright notice and this permission notice shall be included in
|
||||
* all copies or substantial portions of the Software.
|
||||
*
|
||||
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED,
|
||||
* INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
||||
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
||||
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
||||
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
||||
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
||||
*
|
||||
*/
|
||||
template <class T>
|
||||
std::vector<T> get_primes(T max)
|
||||
{
|
||||
std::vector<T> primes;
|
||||
char *sieve;
|
||||
sieve = new char[max / 8 + 1];
|
||||
// Fill sieve with 1
|
||||
memset(sieve, 0xFF, (max / 8 + 1) * sizeof(char));
|
||||
for (T x = 2; x <= max; x++)
|
||||
{
|
||||
if (sieve[x / 8] & (0x01 << (x % 8))) {
|
||||
primes.push_back(x);
|
||||
// Is prime. Mark multiplicates.
|
||||
for (T j = 2 * x; j <= max; j += x)
|
||||
{
|
||||
sieve[j / 8] &= ~(0x01 << (j % 8));
|
||||
}
|
||||
}
|
||||
}
|
||||
delete[] sieve;
|
||||
return primes;
|
||||
}
|
||||
|
||||
//---------------------------------------------------------------
|
||||
//int main() // by Florian Mayer
|
||||
//{g++ -O3 -std=c++14 -fopenmp main.cpp && ./a.out
|
||||
// vector<unsigned long> primes;
|
||||
// primes = get_primes(10000000);
|
||||
// // return 0;
|
||||
// // Print out result.
|
||||
// vector<unsigned long>::iterator it;
|
||||
// for(it=primes.begin(); it < primes.end(); it++)
|
||||
// cout << *it << " ";
|
||||
//
|
||||
// cout << endl;
|
||||
// return 0;
|
||||
//}
|
||||
71
ex1/ex1G_dense_matrices_access/DenseMatrix.h
Normal file
71
ex1/ex1G_dense_matrices_access/DenseMatrix.h
Normal file
|
|
@ -0,0 +1,71 @@
|
|||
#pragma once
|
||||
#include "sigmoid.h"
|
||||
#include <iostream>
|
||||
#include <vector>
|
||||
using namespace std;
|
||||
|
||||
class DenseMatrix
|
||||
{
|
||||
private:
|
||||
vector<double> M;
|
||||
size_t n;
|
||||
size_t m;
|
||||
|
||||
|
||||
public:
|
||||
vector<double> Mult(const vector<double> &x) const
|
||||
{
|
||||
vector<double> y(n,0);
|
||||
for(size_t i = 0; i < n; ++i) // iterate row
|
||||
{
|
||||
for(size_t j = 0; j < m; ++j) // iterate column
|
||||
{
|
||||
y[i] += M[i*m + j]*x[j];
|
||||
}
|
||||
}
|
||||
return y;
|
||||
}
|
||||
|
||||
vector<double> MultT(const vector<double> &y) const
|
||||
{
|
||||
vector<double> x(m,0);
|
||||
for(size_t j = 0; j < m; ++j) // iterate column
|
||||
{
|
||||
for(size_t i = 0; i < n; ++i) // iterate row
|
||||
{
|
||||
x[j] += M[i*m + j]*y[i];
|
||||
}
|
||||
}
|
||||
return x;
|
||||
}
|
||||
|
||||
void Print() const
|
||||
{
|
||||
for(size_t i = 0; i < n; ++i) // iterate row
|
||||
{
|
||||
for(size_t j = 0; j < m; ++j) // iterate column
|
||||
{
|
||||
cout << M[i*m + j] << " ";
|
||||
}
|
||||
cout << endl;
|
||||
}
|
||||
cout << endl;
|
||||
}
|
||||
|
||||
|
||||
DenseMatrix(size_t n, size_t m)
|
||||
{
|
||||
this->n = n;
|
||||
this->m = m;
|
||||
M = vector<double>(n*m);
|
||||
size_t nm = max(n,m);
|
||||
|
||||
for(size_t i = 0; i < n; ++i) // iterate row
|
||||
{
|
||||
for(size_t j = 0; j < m; ++j) // iterate column
|
||||
{
|
||||
M[i*m + j] = sigmoid(x_entry(i,nm))*sigmoid(x_entry(j,nm));
|
||||
}
|
||||
}
|
||||
}
|
||||
};
|
||||
52
ex1/ex1G_dense_matrices_access/ProductMatrix.h
Normal file
52
ex1/ex1G_dense_matrices_access/ProductMatrix.h
Normal file
|
|
@ -0,0 +1,52 @@
|
|||
#pragma once
|
||||
#include <cmath>
|
||||
#include <iostream>
|
||||
#include <vector>
|
||||
using namespace std;
|
||||
|
||||
class ProductMatrix
|
||||
{
|
||||
private:
|
||||
vector<double> u;
|
||||
vector<double> v;
|
||||
size_t n;
|
||||
size_t m;
|
||||
|
||||
public:
|
||||
vector<double> Mult(const vector<double> &x) const
|
||||
{
|
||||
vector<double> y(n,0);
|
||||
for(int i = 0; i < n; ++i)
|
||||
{
|
||||
for(int j = 0; j < m; ++j)
|
||||
{
|
||||
y[i] += v[j]*x[j];
|
||||
}
|
||||
y[i] *= u[i];
|
||||
}
|
||||
return y;
|
||||
}
|
||||
|
||||
vector<double> MultT(const vector<double> &y) const
|
||||
{
|
||||
vector<double> x(m,0);
|
||||
for(int j = 0; j < m; ++j)
|
||||
{
|
||||
for(int i = 0; i < n; ++i)
|
||||
{
|
||||
x[j] += y[i]*u[i];
|
||||
}
|
||||
x[j] *= v[j];
|
||||
}
|
||||
return x;
|
||||
}
|
||||
|
||||
ProductMatrix(const vector<double> &u, const vector<double> &v)
|
||||
{
|
||||
n = u.size();
|
||||
m = v.size();
|
||||
this->u = u;
|
||||
this->v = v;
|
||||
|
||||
}
|
||||
};
|
||||
BIN
ex1/ex1G_dense_matrices_access/main
Executable file
BIN
ex1/ex1G_dense_matrices_access/main
Executable file
Binary file not shown.
93
ex1/ex1G_dense_matrices_access/main.cpp
Normal file
93
ex1/ex1G_dense_matrices_access/main.cpp
Normal file
|
|
@ -0,0 +1,93 @@
|
|||
#include "../utils/timing.h"
|
||||
#include "DenseMatrix.h"
|
||||
#include "ProductMatrix.h"
|
||||
#include <algorithm>
|
||||
|
||||
int main()
|
||||
{
|
||||
// b) ------------------------------------------------------------------------------------------------------
|
||||
DenseMatrix const M(5,3);
|
||||
vector<double> const u{{1, 2, 3}};
|
||||
vector<double> f1 = M.Mult(u);
|
||||
vector<double> const v{{-1, 2, -3, 4, -5}};
|
||||
vector<double> f2 = M.MultT(v);
|
||||
|
||||
M.Print();
|
||||
|
||||
for(size_t i = 0; i < f1.size(); ++i)
|
||||
cout << f1[i] << endl;
|
||||
cout << endl;
|
||||
|
||||
|
||||
for(size_t j = 0; j < f2.size(); ++j)
|
||||
cout << f2[j] << " ";
|
||||
cout << endl << "-------------------------------------------------" << endl;
|
||||
|
||||
// c) ------------------------------------------------------------------------------------------------------
|
||||
size_t n = pow(10,3);
|
||||
DenseMatrix const M_1(n,n);
|
||||
vector<double> x(n, 1.0);
|
||||
|
||||
size_t n_loops = 100;
|
||||
vector<double> y_1;
|
||||
vector<double> y_2;
|
||||
|
||||
double time_1 = 0;
|
||||
double time_2 = 0;
|
||||
|
||||
tic();
|
||||
for(int l = 0; l < n_loops; ++l)
|
||||
y_1 = M_1.Mult(x);
|
||||
time_1 += toc();
|
||||
|
||||
tic();
|
||||
for(int l = 0; l < n_loops; ++l)
|
||||
y_2 = M_1.MultT(x);
|
||||
time_2 += toc();
|
||||
|
||||
vector<double> error_vec(n,0);
|
||||
for(int i = 0; i < n; ++i)
|
||||
error_vec[i] = abs(y_1[i] - y_2[i]);
|
||||
double sup_error = *max_element(error_vec.begin(), error_vec.end());
|
||||
|
||||
|
||||
cout << "n = " << n << endl;
|
||||
cout << "Average duration for Mult: " << time_1/n_loops << endl;
|
||||
cout << "Average duration for MultT: " << time_2/n_loops << endl;
|
||||
cout << "sup-error: " << sup_error << endl;
|
||||
cout << "-------------------------------------------------" << endl;
|
||||
|
||||
// d) ------------------------------------------------------------------------------------------------------
|
||||
vector<double> u_M(n,0);
|
||||
for(int i = 0; i < n; ++i)
|
||||
u_M[i] = sigmoid(x_entry(i, n));
|
||||
|
||||
ProductMatrix const M_2(u_M, u_M);
|
||||
|
||||
time_1 = 0;
|
||||
time_2 = 0;
|
||||
|
||||
tic();
|
||||
for(int l = 0; l < n_loops; ++l)
|
||||
y_1 = M_2.Mult(x);
|
||||
time_1 += toc();
|
||||
|
||||
tic();
|
||||
for(int l = 0; l < n_loops; ++l)
|
||||
y_2 = M_2.MultT(x);
|
||||
time_2 += toc();
|
||||
|
||||
for(int i = 0; i < n; ++i)
|
||||
error_vec[i] = abs(y_1[i] - y_2[i]);
|
||||
sup_error = *max_element(error_vec.begin(), error_vec.end());
|
||||
|
||||
|
||||
cout << "n = " << n << endl;
|
||||
cout << "Average duration for Mult: " << time_1/n_loops << endl;
|
||||
cout << "Average duration for MultT: " << time_2/n_loops << endl;
|
||||
cout << "sup-error: " << sup_error << endl;
|
||||
cout << "-------------------------------------------------" << endl;
|
||||
|
||||
|
||||
return 0;
|
||||
}
|
||||
12
ex1/ex1G_dense_matrices_access/sigmoid.h
Normal file
12
ex1/ex1G_dense_matrices_access/sigmoid.h
Normal file
|
|
@ -0,0 +1,12 @@
|
|||
#pragma once
|
||||
#include <cmath>
|
||||
|
||||
double sigmoid(double x)
|
||||
{
|
||||
return 1./(1. + exp(-x));
|
||||
}
|
||||
|
||||
double x_entry(size_t k, size_t nm)
|
||||
{
|
||||
return (10.*k)/(nm - 1) - 5.;
|
||||
}
|
||||
Loading…
Add table
Add a link
Reference in a new issue