I've tested this only on R2010a, where it passes all my unit tests. (actually, I'll leave it in, because that would turn the answer into a link-only reply, as rightfully remarked by Andy Clifton). I'll post this on the FEX, and after it's approved, I'll place a link to it here.ĮDIT 1 they approved my submission, which can be found here or directly from my GitHub. I will submit a ticket to MathWorks and update the question if I hear back unless someone has another workaround. with the paths added back, n = gpuDeviceCounts still throws the first error). The first is unexpected and could happen for any number of reasons such as a corrupt install or a stupid typo on my part (e.g. The second one is expected and if I could easily and deliberately throw that error, I could see what exception to test for in a try/catch block and behave appropriately. These products offer 'gpuDeviceCount':ĭifferent errors, different exceptions thrown. GpuDeviceCount is not included in your installed products. Undefined function or variable 'gpuDeviceCount'.īut when I asked a friend, who does not have the parallel toolbox installed, to type in the same command, he got the following: > gpuDeviceCount ![]() For instance, after I removed the directories, I get the following behavior: > n = gpuDeviceCount It also throws different errors, and so this method cannot be used to adequately test the behavior in my program. Once that was done, typing in ver still shows that the Parallel Toolbox is installed. First of all, the toolbox is not in a single directory, it is in several (I had to rmpath 11 directories for Parallel Toolbox and I don't really know if I got them all). Simply removing the path does not disable the toolbox. ![]() Is there any way, short of repeatedly uninstalling and then reinstalling the toolbox, or else buying a second license and installing on a second computer (or the same computer with a different username), to temporarily make MATLAB think that it isn't available?įrom the comments to the question, it appears the answer to my question is "no". To generalize the situation, I would like to be able to test the behavior of an application when a toolbox dependency is not met so I can either provide a meaningful message to the user for why they can't use my application or else provide a workaround. The problem is that since I have the toolbox installed, I never get an error when I try to use a parallel toolbox function, so I can't test the failure points. I am writing an application that will be sent out to customers who may not have that toolbox installed and I would like to write my application so that it fails gracefully (i.e. The pictures show some examples of mechanisms designed with the SG-Lib.I have a standalone license for MATLAB R2016b with the Parallel Toolbox. The library grows with discussions and also by code of other researchers: The Toolbox will certainly be further developed until 2035. Until then there is always a new version. The Toolbox license always expires after about one year. The Toolbox can be downloaded at Matlab Central (set of all SG-Lib versions) at GitHub (dedicates SG-Lib versions) for educational non-profit purpose. Tim Lueth, Professor at Technical University of Munich, Germany There is also a connection for processing medical image data from CT/MRT. Likewise, a part of the SG-Lib is dedicated to the simulation of multi-body systems and the automatic generation of control code for the ♜ of the arduino family or the use of 2D or 3D surface cameras. This concerns the construction of rigid bodies, joints, kinematics, gears, but also shape optimization with computer-aided optimization (CAO) or topology optimization (SKO) both for small movements (small displacement) or the mechanism design (large displacement). The aim of the development since 2010 is to automate the design process for the construction of robots. In some cases further Matlab toolboxes are required. ![]() The SG-Lib is a toolbox for the Matlab math program from Mathworks. ![]() Introduction to Solid Geometry Library Toolbox
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |