while loop with shift register

[LvL1995]

I see,

I've implemented an event structure. But clearly I'm not doing it right (I'm not really familiar with it yet) The clipping function doesn't work at all, and when I use a shift register the graph keeps moving when I let it calculate the mean etc. When I don't use shift register the mean function works properly, but then it wouldn't be able to remember the things I clipped (if it works at all).

How can I solve this?

[crossrulz]

1. See those output tunnels with the little dot in them?  That indicates they are set to Use Default If Unwired.  Right-click on them and turn that off.  The only exception here is the Stop condition, that one can default to FALSE for you.

2. Add a shift register for the Max, Min, Average.  You are going to feed through this value in all cases where you do not update it.

3. In your Clipping subVI, you should look into using Array Subset and Replace Array Subset instead of Delete From Array and Insert Into Array.  The Array Subset and Replace Array Subset will avoid memory copies, making things more efficient in both memory and processing.

[LvL1995]

Thanks for your feedback!

I've managed to do what you suggested. Sadly, the graph still moves when I click 'Calculate mean of nr of elements' Do you know how I can stop it from doing that? I've added captures and my new program.

Thank you 🙂

[GerdW]

Hi LvL,

Sadly, the graph still moves when I click 'Calculate mean of nr of elements' Do you know how I can stop it from doing that?

You still repeat this averaging operation on already processed data: no wonder it will average more and more…

Btw. in the event structure you see a big disadvantage of using very long labels (even with a line break inside): the event case is hard to read! Use shorter control labels and use the control caption in the front panel instead!

[LvL1995]

Thanks for the tip on the labels!

I understand the problem, but I really don't know what I have to do to make it work.  I had a program (added it) where it did what I wanted. When I went to value 1 in the 'Calculate mean' it showed the raw data. When I went to 5 it showed the most averaged version of the data. In this program I don't have shift registers. But Crossrulz said I should use shift register In the program with the event structure. I just don't know what's right and what's not anymore.

All I want is to combine the two subVI's in one, working program 😕 

Thanks!

[LvL1995]

Update: I've had a small sudden realisation moment!

The mean sub vi is programmed to handle raw data, not data from the graph! 

So this will be my plan: 

First, the data has to go through the mean subvi. After that it goes into the graph for the first time. When the data is clipped, the data from the graph is my new 'raw data' . From this data I need the new high, low, average and mean. This updated data has to go into the same graph again. The cycle has to repeat with every clip that is done. So this means I can not send the edited graph data through the mean subvi again and again, I have to write something that is suited for the graph data, right? 

Any ideas on this?

[LvL1995]

I'm really starting to get a bit desperate now... I've tried so much but I have the feeling I'm just going in circles with this problem.... Can someone please give some elaborate tips?

[GerdW · Accepted solution]

Hi LvL,

take a sheet of paper.

Write your algorithm definition onto that sheet: which data gets processed? Which data needs to be stored within the loop? When should data be processed (all the time or event-based)?

Then draw the algorithm on that sheet of paper. (Speak with your collegues about your algorithm to verify it.)

Then (and only then) start to code in LabVIEW!

[LvL1995]

Thanks,

Will do that. Maybe start with a clean slate. I'll let you know when I worked it out.

Thanks for all your help so far!

[LvL1995]

Hi GerdW,

It helped! Starting over helped me get it done.

I used a sequence structure so the program runs properly. In the first tab, I read the file and I replace all the zero values with NaN values. Then in the second, I split the 2D array into six 1D arrays. For every array, there is a tab in the sequence with a while loop in which the peaks and valleys can be clipped in the graph. Because of the sequence structure, the user has to edit the graph, then click a button before the next graph can be edited. This feature was unintended but my employer actually liked it. After all the graphs are edited, The arrays are bundled and put into a for loop.  The data is put into a new graph. In the for loop the coating offset, the mean and the Max, Min, and Avg are calculated. 

There are still some minor things to work on, but the program does what I want it to do, and that's the most important for me.

Thanks for all your help!