Hole in Wall - Deleting an Array of Indices from an Array of Indices

[Lavezza]

The very first two elements of "Wall Indices" are the same (3878).

The original code passes both values through since 3878 isn't a "hole" which is why it gives an answer that is 1 larger in size compared to the Map/Set examples. If 3878 was a hole, the original code would only remove the first instance while Map/Set would remove both.

[altenbach]

(Sorry, could not look at your VI, so maybe I misunderstood)

<content deleted>

Edit: Sorry, I misunderstood the problem...

[altenbach]

---

@BertMcMahan wrote:

---

Well heck then, here's a much simpler version:

 

---

Here's an even simpler version.

~20ms in my testing (on a W10 VM where your solution is about 100ms) producing the exactly same result as the original code.

[Yamaeda]

---

@LuminaryKnight wrote:

---

---

So I'm not entirely sure I follow what you're trying to do here.  Let me know if I didn't duplicate it correctly.  But this nearly broke my machine.  Filled up  my memory instantly.

---

That looks correct to me. I just tried with 500k Wall and 10k holes and it was almost instant. It is Array Subset you're using, right? (The picture i posted was a Snippet, if you save it to disk you should be able to drag it into a VI and get the code)

[Yamaeda]

Version 2, with Maps ... i just realized it's basically identical to Altenbachs ... i guess that's a good thing. 🙂

[altenbach]

---

@Yamaeda wrote:
That looks correct to me. I just tried with 500k Wall and 10k holes and it was almost instant.

---

It does not preserve the original order of the elements. You also need to be careful if the input is a diagram constant, because the entire thing could potentially be constant folded once debugging is disabled, making any benchmarking meaningless.

Whenever I hear "almost instant", I am reminded of the comic Steven Wright who once said: "I put instant coffee in a microwave oven and almost went back in time" (The joke made sense back in the time where these ovens were touted as being so much faster than regular cooking.)

---

@Yamaeda wrote:

Version 2, with Maps ... i just realized it's basically identical to Altenbachs ... i guess that's a good thing. 🙂

---

I was using a Set, not a Map (small difference)! 😄

[LuminaryKnight]

---

@altenbach wrote:

---

@BertMcMahan wrote:

---

Well heck then, here's a much simpler version:

 

---

Here's an even simpler version.

~20ms in my testing (on a W10 VM where your solution is about 100ms) producing the exactly same result as the original code.

---

Greatly appreciated.  I did not even know these functions were available.  Adding new stuff to my bag of tricks.  Thank you everyone!

[altenbach]

---

@LuminaryKnight wrote:
Greatly appreciated.  I did not even know these functions were available.  Adding new stuff to my bag of tricks.  Thank you everyone!

---

Sets and Maps are extremely useful and relatively new (introduced in LabVIEW 2019 and probably first suggested in 2010 😄 ).

For an overview, check my presentation.