Labview 8.5 Slow

[crimolvic]

Yes, but I still don't think the application is the issue. Like I said before, the vi ran a lot much faster when I had LV8.2

[Matthew_Kelton]

Well, if you have that VI, then post it.  As I said, I took your 8.5 VI, and saved it for 8.2 and it ran in about the same time in 8.2.  If you don't have your original VI, no one is going to be able to confirm or explain the behavior you say you see.  The only way I see that the VI you posted could have run that fast is if your deltaTethaMáx value was bigger, resulting in less interpolations.

[crimolvic]

I understand what you said, unfortunatelly I can't do that because is the same vi. I made that vi in LV8.2, then I installed LV8,5, I started the vi and it ran much slower.

If I'd knew that the vi would run slower I have maybe made a copy of it and maybe I would'nt have installed LV8.5 on my computer.

Now I can only think on saving all the vi's I opened for LV8.2, uninstall LV8.5 and install LV8.2. However, if you say you have both versions and you don't see any difference, then that would also not help.

It is true, now I can't prove it. Before I ran the vi with an 8s signal sampled at 51,2 kS/s where deltaTethaMáx = 3,22725E-5 and number of interpolations = 411684. The complete process took about 5 seconds.

[altenbach]

---

Matthew Kelton wrote:
I am hoping Altenbach will have a different approach which will speed things up.

---

One possibility would be to simply use the built-in interpolation routines. Not really faster though.

Here's a quick draft....

[Karunya_R]

Hi,

It is rather unusual that your VIs are running slower. I did try opening up your code but it appeared broken. Here is some information relating to built in LabVIEW interpolation function. Like Matthew K stated, the speed of your execution is dependant on the speed of your computer. Can you post VI you are using now and the sub vis associated with it? I can try to look into it (also it would help if the documentation for the VI was in english). How long does it take you now for the interpolations to be completed?

[altenbach]

OK, here's a way to do it in 80ms instead of 8000ms!!!

Simply set the interpolation to "spline" instead if "linear". Don't ask me why, because I don't understand it yet. 😮 😉

Message Edited by altenbach on 02-15-2008 12:07 PM

[Karunya_R]

It might be because linear interpolation uses data fitting and Spline doesn't. Here some information on Spline Interpolation.

[altenbach]

---

Karunya_R wrote:
It might be because linear interpolation uses data fitting and Spline doesn't. Here some information on Spline Interpolation.

---

Since when does linear interpolation do any data fitting??? It's just simple geometry! 🙂

Can you explain?

[Karunya_R]

Sorry for the unclear previous post :

What I meant was that linear interpolation uses linear functions and Spline uses polynomial instead. Questions?

[altenbach]

Yes, but a linear function is just a special polynomial (actually one of the simplest polynomial of all)! Why should it be 100x slower? 🙂