桂田 祐史
Date: 2015年9月1日, 2018年3月1日
という形をしているというのが良く分からなかった。 常微分方程式でないので、
というもの。 ところで
が見えてきて (多分証明出来る)、「あれれ」となる。
c=1; L=1; u[n_, x_, t_] := Sin[n Pi x/L] Cos[c n Pi t/L] Animate[Table[ Plot[{u[n, x, 0], u[n, x, t]}, {x, 0, L}, PlotRange -> {-1, 1}], {n, 1, 3}], {t, 0, 2L/c, 0.01}] |
![]() ![]() ![]() ![]() | ||
![]() ![]() ![]() | ||
![]() ![]() ![]() |
だろうか?
Cauchy-Kowalevski の定理があるな。
双曲型方程式の初期値問題の一意性は何か解析性と関係があるか? 特異性の伝播とか。
楕円型偏微分方程式の解は一般に実解析的なのか?
楕円型方程式の一意接続定理というのがあった。どんなの?
unzip INTLAB.zip cp -pr Intlab_V9 ~/Documents/MATLAB |
cd Intlab_V9 startintlab |
VVVVVVVVVVVVVVVV Installation of INTLAB VVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVV This is the first call of INTLAB, and the best way to switch the rounding mode will be detected. It is a very rigorous check and takes a short moment. This is done only once with the very first call of INTLAB. To find the best routine, various routines are checked. Unfortunately, in very rare cases it might happen that a hard error occurs. This should never happen, but it happes in some Matlab versions and/or operating systems, and it is beyond my control. In such a rare case of a hard error, please restart INTLAB. The circumstances will be stored, and this switching routine will not be called again. Sorry for possible inconveniences. Press Enter to continue. |
~/Documents/MATLAB/startup.m に書き加える |
addpath('~/Documents/MATLAB/Intlab_V9') |
>> startintlab |
tar xzf OpenBLAS-0.2.14.tar.gz cd OpenBLAS-0.2.14 vi Makefile.rule |
make>&make.log sudo make PREFIX=/usr/local install |
cd /opt/local/var/macports/sources/rsync.macports.org/release/tarballs/ports/ cd math/octave/work/octave-3.8.2 |
sudo port install gcc5 |
If your CPU is Sandy Bridge, please use Clang version 3.1 and above. The Clang 3.0 will generate the wrong AVX binary code of OpenBLAS.えー?Core i7 というのはそれに入るのか? Clang のバージョンばかり気にしているようだけど、 GCC は使うなということなのかな。
tar xzf OpenBLAS-0.2.14.tar.gz cd OpenBLAS-0.2.14 make>&make.log |
sudo make PREFIX=/usr/local install |
# CONSISTENT_FPCSR = 1 |
make CC=gcc-mp-5 FC=gfortran-mp-5 |
Cannot find plugin 'liblto_plugin.so' |
ln -s /opt/local/bin/ar gcc-ar ln -s /usr/bin/ar gcc-ar |
tar xzf OpenBLAS-0.2.14.tar.gz cd OpenBLAS-0.2.14 vi Makefile.rule make>&make.log sudo make PREFIX=/usr/local install |
sudo port install openblas C-z (中断) pushd /opt/local/var/macports/sources/rsync.macports.org/release/tarballs/ cd ports/math/OpenBLAS/work/OpenBLAS-0.2.14 vi Makefile.rule(CONSISTENT_FPCSR = 1 を有効にする。) popd fg |
tar xzf OpenBLAS-0.2.14.tar.gz cd OpenBLAS-0.2.14 vi Makefile.rule make CC=gcc FC=gfortran>&make.log sudo make PREFIX=/usr/local install |
LAPACKから lapack-3.5.0.tgz を入手する。
tar xzf lapack-3.5.0.tgz cd lapack-3.5.0 cp INSTALL/make.inc.gfortran make.inc vi make.inc(特に書き換えるべきところはない? 以前は PLATFORM を LINUX から DARWIN にしたりしたけど。) make blaslib variants>&make-blaslib-variants.log make >&make.log |
sudo cp -p liblapack.a /usr/local/lib/ sudo cp -p libtmglib.a /usr/local/lib/ |
cd lapack-3.5.0 make lapackelib |
sudo cp -p liblapacke.a librefblas.a /usr/local/lib sudo cp -p lapacke/include/*.h /usr/local/include |
gfortran -I/usr/local/include test-lapacke.c \ -L/usr/local/lib -lmatrix -llapacke -llapack -lopenblas |
% ./a.out A= 11 12 13 0 0 0 0 0 21 22 23 24 0 0 0 0 31 32 33 34 35 0 0 0 0 42 43 44 45 46 0 0 0 0 53 54 55 56 57 0 0 0 0 64 65 66 67 68 0 0 0 0 75 76 77 78 0 0 0 0 0 86 87 88 b= 74 230 505 890 1385 1990 1994 1829 call dgbsv successful return from dgbsv 計算して得た解= 1 2 3 4 5 6 7 8 結果が 1,2,3,.. となっていたら成功 % |
port variants octave |
sudo port install octave +accelerate +gcc5 |
Octave は BLAS をダイナミック・リンクするので、 コンパイルし直す必要がないという説もある。
cd /opt/local/var/macports/build/ cd _opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_math_octave cd octave/work/octave-3.8.2 sudo make sudo port install octave +accelerate +gcc5 |
sudo port install octave +accelerate +gcc5 +x11 |
otools -L /usr/local/octave/3.8.0/bin/octave-3.8.0 |
libinterp/corefcn/dlmread.cc-tst ....................... PASS 12/20 FAIL 8 libinterp/corefcn/sparse-xpow.cc-tst ................... PASS 1/2 FAIL 1 libinterp/corefcn/str2double.cc-tst .................... PASS 28/31 FAIL 3 scripts/io/importdata.m ................................ PASS 22/23 FAIL 1 scripts/sparse/svds.m .................................. PASS 5/6 FAIL 1 scripts/specfun/realpow.m .............................. PASS 4/5 FAIL 1 io.tst ................................................. PASS 88/91 FAIL 3 |
The HDF5 header files used to compile this application do not match the version used by the HDF5 library to which this application is linked. Data corruption or segmentation faults may occur if the application continues. This can happen when an application was compiled by one version of HDF5 but linked with a different version of static or shared HDF5 library. You should recompile the application or check your shared library related settings such as 'LD_LIBRARY_PATH'. You can, at your own risk, disable this warning by setting the environment variable 'HDF5_DISABLE_VERSION_CHECK' to a value of '1'. Setting it to 2 or higher will suppress the warning messages totally. Headers are 1.8.15, library is 1.8.14 |
General Information: ------------------- HDF5 Version: 1.8.14 Configured on: Tue Jan 13 07:47:08 CET 2015 Configured by: hecht@MBP-FH.local Configure mode: production Host system: i386-apple-darwin14.0.0 Uname information: Darwin MBP-FH.local 14.0.0 Darwin Kernel Version 14.0.0: Fri Sep 19 00:26:44 PDT 2014; root:xnu-2782.1.97~2/RELEASE_X86_64 x86_64 Byte sex: little-endian Libraries: static, shared Installation point: /usr/local |
-with-hdf5-includedir=/opt/local/include -with-fftw3-includedir= -with-fftw3-libdir
./configure -prefix=/opt/local -includedir=/opt/local/include -libdir=/opt/local/lib
sudo port install octave +accelerate +gcc5 +glgui |