I'm trying to load an Armadillo binary file with Armadillo mat::load() function.
On Windows it worked fine.
I recently converted my code to Linux and now it gives me the error: Mat::load(): couldn't read <filename>.
The file exist in that location and I have read permissions.
What else can be the problem?
Thanks.
Loading and saving binary files in Armadillo is currently platform specific. In other words, if you save in Windows, you may not be able to load the file in Linux. However, if you save in Linux, it will also load in Linux.
The problem was that I had a trailing newline char in my filename string.
In windows it didn't matter, but Linux is picky.
Related
I've been trying to load a library into lua file. Sparing the details, as they are not really important, I have tried this many ways.
The final way, and the one I believe to be correct although I still can't get it to work, is to use "package.loadlib". See code:
ed = package.loadlib("Encode_Decode.lua", "luaopen_ed")
print(ed)
But when I run the program I get this error:
Encode_Decode.lua is either not designed to run on Windows or it
contains an error. Try installing the program again using the original
installation media or contact your system administrator or the
software vendor for support.
I know the program runs because I used it internally to test it's encoding and decoding abilities and it worked fine. I'd really prefer not moving the contents of the library over as my main lua file is crowded as it is. I will if I have to though.
Yes it is in the main folder. I've also tried changing the extension of the library file into a .dll, with the same error.
What am I doing wrong?
I apologize in advance if this is a duplicate, I did my best to research this problem as thoroughly as I could. But to be honest it's almost 3 AM and I've been searching for almost an hour.
Stupid beginner mistake, used the wrong syntax.
require("Encode_Decode")
print(dec("bnVs")) --returns "nul"
package.loadlib is used for loading shared libraries; i.e. .dll or .so files. You're passing a .lua file to it, so Windows attempts to load it as a .dll and fails when it can't.
To load Lua source code, you can use dofile. Alternatively, you can use require, which is a bit more complex, but handles loading modules only once and works with both Lua and C modules.
I have a map file created by delphi compiler and I need to use it to analyze DMP file.
I converted the map file to dbg file, using Map2dbg.
The problem is that the process has a different signeture then the symbols.
I tried to use chkmatch to match the signeture, but I'm getting:
Error: Debug information not found in the executable
I guess this is because it is a delphi process.
Does anyone know how to match the signatures?
I made a minidumpreader some time ago:
https://asmprofiler.googlecode.com/svn/trunk/MiniDumpReader/ViewMinidump.exe (old link)
https://github.com/andremussche/asmprofiler/tree/master/Releases/ViewMinidump.exe (new link)
It is written in Delphi with jclDebug.pas so it supports all delphi debug symbols.
And because of this it has line number support (which map2dbg or tds2pdb don't have).
Note: I haven't made map2dbg or tds2pdb myself, just hosting it so other can easily find it
I need some help converting a cpp file to a mex file.
I get the error below:
fatal error: opencv2/core/core.hpp: No such file or directory
Compilation terminated
The cpp file is using opencv header files. I'm not sure how to link Matlab with opencv in Ubuntu. I saw this link
http://xanthippi.ceid.upatras.gr/people/evangelidis/matlab_opencv/
but its for Windows and I'm not sure where Ubuntu stores opencv. There are a couple places I found opencv folders
usr/local/include/opencv and
usr/local/include/opencv2 and
usr/local/share/OpenCv
Tried these with OCVROOT but got 'Error: Unexpected Matlab Operator'. Not sure what I'm doing wrong here. Any help would be much appreciated.
Thank you!
Mex needs to be told where the OpenCV header files are. You can either fix it in ~/.matlab/R2013a/mexopts.sh or just put a -I argument on the mex command line. You will find the answer to How to link during Matlab's MEX compilation helpful.
You will have to sort out which of those OpenCV versions you want to use. Might be best to install you own so you know what you are dealing with.
'Error: Unexpected Matlab Operator' would be cause by an error in your matlab code.
matlab_opencv should work fine on Linux, the only windows specific thing about it the instructions, as far as I know.
unless you have a custom mexopts.sh, I don't think setting OCVROOT will hve any affect on Matlab.
I am completely new to CUDA and C. My eventual goal is to adapt a FORTRAN program for CUDA. For now, I am having trouble getting a simple "Hello world" program to run.
I have been able to compile using nvcc hello_test.cu --machine32 which creates an executable. However when I try to run the .exe, I get a System Error message stating:
The program can't start because cudart32_32_9.dll is missing from your computer. Try reinstalling the program to fix this problem
I imagine I haven't added the path to this file (which I can find in "C:\ProgramData\NVIDIA Corporation\NVIDIA GPU Computing SDK 4.2\C\common\bin\"); but I don't know how to add the path. Any help would be appreciated.
You can go several routes. Either just make a copy of the DLL and put it in the same directory as your executable.
Or you can add the path you mentioned in the PATH environment variable of Windows. Then it should be able to find the DLL.
Okey, this is not a core programming question; it is more of a question regarding cgns (CFD general notational system) API.
I've exported a grid/mesh file from ANSYS Fluent (which was first created in Gambit 2.46), and I wrote a very simple Fortran program to open and close it (doing nothing else). To check the file is not corrupt I plotted it in Tecplot.
So, when I compiled using gfortran with the mentioned cgns and ran the program I got this error (as part of cg_error_exit_f())
ADF_Database_Open:File does not exist or is not a HDF5 file
Here is the program
program cavity
include "/usr/include/cgnslib_f.h"
call cg_open_f("Cavity.cgns",CG_MODE_READ,index_file,ier)
!check for error if so exit
if (ier .ne. CG_OK) then
call cg_error_exit_f()
end if
write(*,*)"I kind of opened the file?"
call cg_close_f(index_file,ier)
stop
end program cavity
I'm able to write both structured and unstructured grids in cgns format, without any problem.
I suspect the cgns library I'm using(version 2.5.5 packaged in Fedora 15 and Scientific linux 6.1) is built to support only HDF5, while the exported grid file is written in ADF format.
Any ideas to circumvent this or perhaps adding ADF? Which by the way is not packaged in both the distributions. Any other grid generator which is compatible with cgns version 2.5.5?
I hope I was clear. Any further info required, I would provide.
There is so much that could've gone wrong in here, and I'm afraid you didn't exactly narrow the problem down.
You said you exported a file from Fluent (what kind of a file is it? Be sure!). cg_error_exit_f() gave you an error listed. I'm assuming you have the source of the mentioned routines? In the program you include a cgnslib_f.h file - what's in it? I'm assumming the program compiled without errors of any kind, making this a file format question, not a fortran question.
Again, verify what kind of file Fluent produced.
When I ran into this situation, I discovered the following tools:
hdf2adf
adf2hdf
They are in the cgns-convert package on Ubuntu and are probably available for your distribution as well.