Labview help with arrays, clusters, and variants

[GastonBA]

Hello,

Need some help, there are too many things that I do not know where to start.

I have a multichannel system (meaning one measuring system with one, lets say, a power supply, a switch, many resistors, and one multimeter).

I want to make all the channels sequencial measurements with the switch going by all channels (assume 8).

Instead of having a work area of the 8 channels repeating the same thing, I want to bundle stuff (parameters, some reference measurement for some later calculation), and select them to send to a single channel measuring step.

In the picture I am showing the channels (in this example 8), and for the DUT name, an array with the string names. I could select the name of the sample with the Index Array if I place it on the for loop (same for other 1 dimensional parameters).

But assume now that I want to measure an V-I curve for a reference resistor (to normalize measurements or correct for different switching channels). I will be taking a set of values that I will be using to normalize the data. That is I have a vector of n points length.

How I do the same thing for the nx1 arrays that I want to bundle and then selectc?

Thank you very much,

Gaston

 

[RavensFan]

Bundle each array into its own cluster.  Then Build an Array of those clusters.

[altenbach]

• Are you sure an I64 datatype to hold the number of channels is reasonable? (Do you really have more than a few billion channels??? Did you notice the coercion dot? the FOR loop expects I32 for N)
• Index array with the index wired to [i] is about the same as autoindexing. (with some important exceptions).
• Sorry, I did not get your real question. Too confusing! Can you attach some code (instead of a picture) and use the control labels in your problem description (e.g. what are currently "string names", which one is "vector of N points", where is the "nx1 array"? etc.))
• Your subject mentions "variants". Where are they?

[GastonBA]

1. Integer, if is I64 does not make a difference if is i8;

2. I been using Labview since 4.0 version where everything use to be wired. Then, last 10 years have not used it. So I may not be using all the new stuff available, like the autoindexing;

3. The code is what I show, literally;

4. See picture from Labview help. I got there from some help topics that lead me there.

 

[altenbach]

1. It is best to use the native datatype, which is I32 for N. No need for glaring coercion dots.
2. I've been using LabVIEW since 4.0 and auto-indexing was available. (In fact autoindexing was there from the very beginning, see the LabVIEW 1.2 manual :D)
3. A picture is never "code" because we cannot see the important details (e.g. the values and the lengths of the various arrays, etc.)
4. Yes, but your post never mentions variants. I could not see the connection. Are you looking for a variant based solution?

I am willing to help, but there is insufficient information to even try....

[GastonBA]

Hello,

1. I am a physicist using Labview to capture data, not programming Labview full time, so I do not have gone that deep. The I64 was default, I did not select it nor correct it;

2. Never saw it before. I am aware now;

3. The important details are irrelevant in this case. The arrays today are 2^11 points long (2048 measurement point, until I find the time to trim them), with DBL or EXT data representation.

4. Variants are in the title, because browsing a solution to my problem I stumble on this category that I never used before and the help is useless. I am not looking for a Variant based solution, I just want to do what I shown for the strings (that is, run a bunch i.e. 8, of measurements in sequential fashion, and use a reference to normalize data).

[altenbach]

Build the ref arrays into a 2D array and autoindex at the loop boundary. You get one 1D array per iteration.

I assume the string array also has 8 elements. 

Autoindexing will loop until it runs out of elements, no need to wire N at all.

There is no A/D converter with so many bits that you require EXT. That just slows you down.