Request for LabVIEW Block Diagrams for Dataset Collection

[Elbek Keskinoglu]

Dear LabVIEW users,

 

I am currently working on a project to create an API that can detect block diagrams of experimental setups in LabVIEW using machine learning. To do this, I need to collect a dataset of LabVIEW block diagrams that represent different experimental setups.

 

I am reaching out to the LabVIEW community to ask for your help in collecting this dataset. If you have any LabVIEW block diagrams that represent experimental setups that you would be willing to share, please send them my way.

 

All block diagrams will be used for research purposes only and will be kept confidential. I will provide credit to contributors in any publications or presentations resulting from this work.

 

Thank you in advance for your help in this project. Please let me know if you have any questions or concerns.

 

Best regards,

Elbek Javokhir Keskinoglu

[ShaunR]

You'll need to define exactly what "experimental setups" means - I've no idea

[Elbek Keskinoglu]

I want any setup. No matter. I want to collect a dataset of LabVIEW block diagrams that represent different experimental setups. This dataset will be used to train a machine learning model to detect the block diagrams, and the data and the experimental setup will be used to recommend how the investigation can continue from the similar setups used by articles and patents. Right now, I am trying to diversify the dataset to annotate the block diagrams. After annotating the dataset, I will train a machine learning model to detect the block diagrams.

[ShaunR]

37 minutes ago, Elbek Keskinoglu said:

represent different experimental setups

I've really no idea what this means.

If you are just looking for a volume of any code then VIPM has hundreds of libraries (with source) that you can train on. You will have 10's of thousand of VI's with source to point your algorithm at. Apart from that, you will have to be more specific.

Artificial Insemination is coming for us, boys and girls.

Edited March 4, 2023 by ShaunR

[jcarmody]

This sounds ambitious.

>> After annotating the dataset, I will train a machine learning model to detect the block diagrams.
This sounds like it should come first?  This feels more approachable because, as Shaun said, there is a lot of code available in VIPM.  There's also a ton on these (and NI's) fora that can easily be labeled as having a block diagram.  Labeling a diagram as an "experimental setup" is more difficult; labeling it with the specific type of setup will be even harder and judging the quality of the programming (so you can "recommend how the investigation can continue from the similar setups") will be (nearly) impossible.

Good luck, though.

[jcarmody]

Cross-posted on NI's forum.

[Phillip Brooks]

Good luck with that...

[Elbek Keskinoglu]

I found a solution, but gotta experiment

[Elbek Keskinoglu]

prototype#1.viThis is what I came up with. Can you let me know how I can enhance it. Here purpose was to receive every data from the input cluster and transform it to JSON. I could simplify as much as this.

[jcarmody]

I have latest LabVIEW 2020.

[Elbek Keskinoglu]

Here it is

[Neil Pate]

@Elbek Keskinoglu are you trying to get a textual representation of a data structure or the block diagram itself?

LabVIEW does actually have a way to save a VI as some kind of XML file or something, but its a bit hidden away and I don't actually know how to make this work. Others in the forum probably will though. See here if you are curious.

Edited March 8, 2023 by Neil Pate

[Elbek Keskinoglu]

At first I wanted to do this through understanding the block diagram. But now, I am doing it by simply putting the block diagram to the input cluster and collect the graph and similar data in the json format.

[Neil Pate]

5 hours ago, Elbek Keskinoglu said:

At first I wanted to do this through understanding the block diagram. But now, I am doing it by simply putting the block diagram to the input cluster and collect the graph and similar data in the json format.

Sorry I do not follow the you. The block diagram is the code, it is not represented by a cluster that you can wire in to your json writing code. 

[ShaunR]

I think you need to do some LabVIEW courses and/or tutorials.

You're clearly are out of your depth here with even the concept of what LabVIEW is and how it operates at the most basic level - you have bitten off more than you can chew.

Go and ask ChatGPT how to make a LabVIEW program to add two numbers and you will see the difficulties you face. 

[Elbek Keskinoglu]

@Niel Pate sorry for my misexplaining. I tried to mean that I am putting the experimental setup at the Front Panel inside of the Cluster. So when there are graphs and charts they will be transformed to the JSON format to be saved in a file.

[ensegre]

48 minutes ago, Elbek Keskinoglu said:

I tried to mean that I am putting the experimental setup at the Front Panel inside of the Cluster. So when there are graphs and charts they will be transformed to the JSON format to be saved in a file.

This text has been generated by AI, right? Because it reads like that BS...🤣

We're getting old. In my time, we were just happy laughing about papers generated by "context free grammar generators".

[Neil Pate]

1 hour ago, Elbek Keskinoglu said:

@Niel Pate sorry for my misexplaining. I tried to mean that I am putting the experimental setup at the Front Panel inside of the Cluster. So when there are graphs and charts they will be transformed to the JSON format to be saved in a file.

I think you need to share some code of what you are trying to do, as this still does not really make sense to me. When you say "experimental setup" I think you mean some kind of configuration controls on the FP of a VI, but I do not see how that related to some graphs and charts unless you look at the code also.