Origin Ver. 2016 Pro and Service Release SR0 (Select Help-->About Origin): Operating System: Windows 7 Enterprise.
Having upgraded to Origin 2016 from 2015 I am now having problems with very large datasets.
I have datasets in the order of 4 million points (4 seconds of data at 1,000,000 samples/sec). My *.opj files can reach several GB in size.
In Origin 2015 I could run the x-functions commands or use the analysis menu to process my datasets and it worked fine. eg: mdwt ix:=col(3) type:=%s$ ext:=pad0 level:=$(extralevels) rd:=[bkname$]MDWT!;
Now in 2016 this command fails (with this size dataset) if I reduce the dataset to say 400,000 points (x10 decimation) it works. It does not matter if I run the code or used the MDWT from the analysis menu.
Below is an capture of a failed MDWT ran from the analysis menu. interesting I believe the higher order columns are correct (when comparing to a series of dwt runs) its just the lower order columns are non-existent. I can successfully perform a single discrete wavelet transform (dwt)on my 4 million point dataset using both code or menu.
A second problem I have found is performing a FFT on these same datasets. produce zero results see attached image.
Again the FFT worked on the smaller dataset.
I repeat, I did not have this problem with Origin 2015.
I do not know if other analysis menu items are affected.
I checked with the SR 1&2 on the support page and no mention of a fix for this was mentioned.
Problems with very large datasets with wavelets
Posted - 05/26/2016 : 03:11:34 AM
