UFF 58 handling

[sbtlaarzc]

The software currently supports only uff58 format, when time/frequency spacing is set to uneven (data in 3 columns in the file). If the spacing is even (data in 4 columns) the file cannot be imported. The former would be a nice improvement, in order to allow the import of files from other software, which use the "even spacing" format.

[NicolasBaleynaud]

Hello, I tried to upload measurement data out of a matlab created uff-data, and I think I couldn't because of this issue. My question is, if you might have found a way to use also even spacing uff datas?

Thank you in advance! Nicolas Baleynaud

[jankoslavic]

Dear Nicolas. If you send me your file I will try to resolve it.

[NicolasBaleynaud]

Hello,

Thank you for your fast answer. The file was actually just the frequency response of a one mass oscillator to test how the software works. Actually github doesn't seem to support the unv type, I saved it as text file, like Silvano Pecaro said: just replace the .txt by .unv .

1m.txt

Thank you in advance!

[jankoslavic]

Dear Nicolas!

The data that you send looks fine and is read without problems with the last version of pyuff. PS: the data is even spaced, this means that the frequency axis ("x") is defined in your case with 1.00000e+00 5.69430e-02 (freq start, freq increment)

due to this four fields per line can be used to store the data. Here is a Fig of the data:

If the freq axis is uneven (e.g. 0Hz, 1Hz, 5Hz,...) then the first column is used for freq axis and the three columns can be used for the data.

@sbtlaarzc can you please check if you are able to import the data into OpenModal?

[jankoslavic]

dear Nicolas!

@sbtlaarzc double-checked your file. you have three datasets (I was looking only at the first) and the last two are wrong/empty (dataset starts and ends with -1)

be careful: if you save to an existing uff file the data can be appended

[NicolasBaleynaud]

Dear Janko

Thank you very much for your help! I didn't get everything, i have saved a complex transfer function as measured data with even spacing. The first column and third column are the real part of th transfer function answer and the second and fourth are the imaginary part. Or do you mean something else by dataset ?

I attached you the matlab program and the data which I used to create the file.

Thank you in advance

uff.zip

[jankoslavic]

Dear Nicolas!

Try removing everything after the second "-1" line in the uff file (this is some 20 lines at the end)

[jankoslavic]

Dear @NicolasBaleynaud ! I was now able to check your zip file; I guess you run the writeuff('1m.unv',unvdata); line once with unvdata containing data and after this twice with unvdata=empty.

[NicolasBaleynaud]

Dear Janko, thank you for taking so much time on this. I guess I did like you said but this was a mistake. after reading your comment on the openadaptronikblog I guess it must be a problem of environment. It seems there is data in Openmodal but I cannot really see it. In the measurement tab above the acceleration there is written Overload! Double hit!, and I didn't pay attention to it before there is a libpng warning: iCCP: known incorrect sRGB profile in the command window, so I guess it must be the reason why I don't see any data.

I just thought that it might be because´of the issue #14 that openmodal cannot handle uff58data with even spacing in four columns, same data type as I have. But you say it should work ?

[jankoslavic]

Hi @NicolasBaleynaud ! Thank you for the message; double hit and overload only make sense when you prepare measurements with OpenModal; we will prepare a short video on how to import uff and combine it with geometry. Once the video is available, you will get the link here.

[jankoslavic]

There is a bug in the exe version (https://github.com/openmodal/OpenModal/issues/28). We need to resolve it first :( Sorry, for this!

[sbtlaarzc]

Hey,

the video now available here: https://youtu.be/avOTGbkvzS8

Note, that this was done by running the code on my branch (blaz).

The exe version currently does not support UFF handling.

[jankoslavic]

Great @sbtlaarzc

[NicolasBaleynaud]

Thank you! I will try it with your video! Nicolas

[jankoslavic]

Dear @NicolasBaleynaud ! If you work with source files, I would advise you to try out the ubuntu-linux-tested branch which works also on Windows (you can work with the Anaconda distribution and add the required packages)

[jankoslavic]

Dear @NicolasBaleynaud if you work with source files, I would advise you to check out the branch ubuntu-linux-tested which works with latest Anaconda + the additional packages which are given in the requirements.txt

[moritzhuelsebrock]

Dear @jankoslavic, I'm Nicolas's successor. I have some problems running the source files. I installed the latest version of Anaconda and added all the required packages. I was using the ubuntu-linux-tested branch, as you suggested in the previous post. The program starts and seems to work just fine but when I start to upload the data or go to Geometry, the program crashes everytime. Do you have an idea, where I could start looking for mistakes?

I also tried to run the blaz branch @sbtlaarzc too . When I try to run the code in the virtual-environment, I get an error text like this: (virtual-environment) (C:\Users\morit\Anaconda3) C:\Users\morit\Documents\Masterarbeit\OpenModal\OpenModal-blaz\OpenModal>python openmodal.py Traceback (most recent call last): File "openmodal.py", line 27, in from PyQt4 import QtGui ModuleNotFoundError: No module named 'PyQt4' I checked and PyQt4 is installed... Do you maybe have an idea where I could start looking for mistakes in my procedure?

Thank you in advance!! Moritz Hülsebrock

[sbtlaarzc]

I am not sure what would be the reason. Do you have both PyQt4 and PyQt5 installed? Maybe there are some compatibility issues.

[moritzhuelsebrock]

@sbtlaarzc thank you for your quick response!

At the moment both versions are installed on my computer. Which branch do you recommend to use for windows?

[matmr]

Hi @moritzhuelsebrock

Your error seems odd. There should be no references to PyQt4 in ubuntu-linux-tested branch. I double-checked just now. In fact, the 27th line of openmodal.py, where your error occurs, reads:

from PyQt5 import QtGui, QtWidgets, QtWebEngineWidgets

Could you check this specific line in your local openmodal code? So we are sure you have the right source.

[moritzhuelsebrock]

Hi @matmr , I think I opened it in my normal python environment and not in a virtual one... I can open it now and it works somehow. There are still some things that don't work, e.g. everytime I click on the "Geometry" or "Geometry" button the program crashes. I get following error message:

OpenGL.error.GLError: GLError( err = 1282, description = b'Der Vorgang ist ung\xfcltig.', (The process is not valid) baseOperation = glViewport, cArguments = (0, 0, 1702, 882) ) Do you might know where this could come from?

Thank you in adcvance! Moritz

[matmr]

@moritzhuelsebrock !

I was not around much past month. Sorry for the late response.

Couldn't say at the moment what's causing the error. Again it is kind of unexpected that the change of the Python environment would cause the previous error. Not sure.

With regards to the geometry problem I'd ask @miha-pirnat He could have an idea about what's happening there.

I hope you're enjoying the holidays :)

mm

[jankoslavic]

@sbtlaarzc started from a different branch... not the master...

[sabhyank1]

Very nice piece of software ! Works flawlessly the first time I installed it.