Re: [arts-users] cloudy simulations with ARTS

2019-12-03 Thread Oliver Lemke 
Hi Paolo,

For blas and lapack: Please make sure you have the libraries and development 
header files installed. I think the packages you have to install with yum on 
CentOS are called lapack-devel and libblas-devel. Having those hopefully fixes 
the compilation with GCC.

Since we use GCC and LLVM for development, the Intel Compiler is currently not 
tested for the development version. I recommend to stick with GCC for now.

Cheers,
Oliver


> On 3. Dec 2019, at 20:07, PAOLO VEGLIO  wrote:
> 
> Patrick,
> yes, I got a few complains by arts about variables/methods not being defined 
> if I try to use v2.2.64.
> 
> I was just looking at the development version today, but I have some issues 
> compiling it (on Linux CentOS 7). I made sure that my machine meets all the 
> requirements but I get a couple different errors, depending on whether I’m 
> using gcc or the intel compiler.
> Namely, if I use gcc, cmake can’t find lapack. This is the message I get when 
> I run cmake:
> 
> CMake Error at cmake/modules/FindBLAS.cmake:699 (message):
>   A required library with BLAS API not found.  Please specify library
>   location.
> Call Stack (most recent call first):
>   cmake/modules/FindLAPACK.cmake:162 (find_package)
>   CMakeLists.txt:248 (find_package)
> I’d provide cmake with the path to my lapack but I’m not sure how to do it. 
> Given what follows though, I’m not sure I would be able to compile the code 
> anyway.
> Cmake works if I use the intel compiler, but then the compilation fails with 
> this error:
> 
> [  8%] Built target microhttpd
> Scanning dependencies of target matpack
> [  8%] Building CXX object src/CMakeFiles/matpack.dir/complex.cc.o
> icpc: command line warning #10121: overriding '-std=c++11' with '-std=c++14'
> /home/pveglio/arts/src/complex.cc (1529): error: more 
> than one operator "*" matches these operands:
> function "std::operator*(const std::complex &, const 
> std::complex &)"
> function "operator*(Complex, Complex)"
> operand types are: const Complex * const Complex
> for (; ai != ae; ++ai, ++bi) res += (*ai) * (*bi);
>   ^
> 
> [...]
> 
> compilation aborted for /home/pveglio/arts/src/complex.cc 
>  (code 2)
> make[2]: *** [src/CMakeFiles/matpack.dir/complex.cc.o] Error 2
> make[1]: *** [src/CMakeFiles/matpack.dir/all] Error 2
> make: *** [all] Error 2
> 
> there’s a whole lot in between but I can’t tell if it could be useful or it’s 
> just the same error propagating throughout the code.
> 
> Let me know if you have any idea of what could cause these errors or you need 
> more information.
> Thanks,
> 
> Paolo
> 
> 
> Paolo Veglio, Ph.D.
> Space Science and Engineering Center
> University of Wisconsin - Madison
> 1225 W. Dayton St.
> Madison WI 53706, USA
> 
>> On Dec 3, 2019, at 12:33 PM, Patrick Eriksson > > wrote:
>> 
>> Paolo,
>> 
>>> And yes, these controlfiles seem to be available only in the development 
>>> version of ARTS, but I assume I can use them with the latest stable version 
>>> as well.
>> 
>> Sorry, but that is not the case. There are features used in those cfiles 
>> that are not at hand in v2.2.
>> 
>> In general, as you are interested in simulations around ICI, I would suggest 
>> that you switch to the development version. We have added quite a lot around 
>> scattering calculations since v2.2.
>> 
>> Bye,
>> 
>> Patrick
> 
> ___
> arts_users.mi mailing list
> arts_users.mi@lists.uni-hamburg.de
> https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi

___
arts_users.mi mailing list
arts_users.mi@lists.uni-hamburg.de
https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi

Re: [arts-users] cloudy simulations with ARTS

2019-12-03 Thread PAOLO VEGLIO 
Patrick,
yes, I got a few complains by arts about variables/methods not being defined if 
I try to use v2.2.64.

I was just looking at the development version today, but I have some issues 
compiling it (on Linux CentOS 7). I made sure that my machine meets all the 
requirements but I get a couple different errors, depending on whether I’m 
using gcc or the intel compiler.
Namely, if I use gcc, cmake can’t find lapack. This is the message I get when I 
run cmake:

CMake Error at cmake/modules/FindBLAS.cmake:699 (message):
  A required library with BLAS API not found.  Please specify library
  location.
Call Stack (most recent call first):
  cmake/modules/FindLAPACK.cmake:162 (find_package)
  CMakeLists.txt:248 (find_package)
I’d provide cmake with the path to my lapack but I’m not sure how to do it. 
Given what follows though, I’m not sure I would be able to compile the code 
anyway.
Cmake works if I use the intel compiler, but then the compilation fails with 
this error:

[  8%] Built target microhttpd
Scanning dependencies of target matpack
[  8%] Building CXX object src/CMakeFiles/matpack.dir/complex.cc.o
icpc: command line warning #10121: overriding '-std=c++11' with '-std=c++14'
/home/pveglio/arts/src/complex.cc(1529): error: more than 
one operator "*" matches these operands:
function "std::operator*(const std::complex &, const 
std::complex &)"
function "operator*(Complex, Complex)"
operand types are: const Complex * const Complex
for (; ai != ae; ++ai, ++bi) res += (*ai) * (*bi);
  ^

[...]

compilation aborted for /home/pveglio/arts/src/complex.cc 
(code 2)
make[2]: *** [src/CMakeFiles/matpack.dir/complex.cc.o] Error 2
make[1]: *** [src/CMakeFiles/matpack.dir/all] Error 2
make: *** [all] Error 2

there’s a whole lot in between but I can’t tell if it could be useful or it’s 
just the same error propagating throughout the code.

Let me know if you have any idea of what could cause these errors or you need 
more information.
Thanks,

Paolo


Paolo Veglio, Ph.D.
Space Science and Engineering Center
University of Wisconsin - Madison
1225 W. Dayton St.
Madison WI 53706, USA

On Dec 3, 2019, at 12:33 PM, Patrick Eriksson 
mailto:patrick.eriks...@chalmers.se>> wrote:

Paolo,

And yes, these controlfiles seem to be available only in the development 
version of ARTS, but I assume I can use them with the latest stable version as 
well.

Sorry, but that is not the case. There are features used in those cfiles that 
are not at hand in v2.2.

In general, as you are interested in simulations around ICI, I would suggest 
that you switch to the development version. We have added quite a lot around 
scattering calculations since v2.2.

Bye,

Patrick

___
arts_users.mi mailing list
arts_users.mi@lists.uni-hamburg.de
https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi