Scripting Helpers is winding down operations and is now read-only. More info→
Ad
Log in to vote
1

How do you quickly iterate through a table with tens of thousands of records?

Asked by 6 years ago
Edited 6 years ago

Exactly what the title says... I ran into a snag where a for in pairs isn't fast enough to check through tens of thousands of records (only does 60 per second). I need to check to see if a part's specific position is already inside a table so that when I generate another part from the user's input, the part will be placed in an area that has never been occupied.

I couldn't find any more information on something called maps and would appreciate some clarity as it may seem as though it could solve my problem.

Here is some quick code just to help show what I mean:


local myTable = {} local part = game.Workspace.Part local testerPart = game.Workspace.testerPart local testPosition = Vector3.new(10,10,10) local counter = 0 ---- The following while loop is just to show you guys, I know they are all the same values while counter < 1,000 do part:Clone() table.insert(myTable, part.Position) counter = counter + 1 end ---- This part is just too slow for i, v in pairs(myTable) do if v == testPosition then break -- since it's already inside the table elseif i = #myTable then ---if it reached the last record without finding a match testerPart:Clone() ----then copy the testerPart testerPart.Parent = game.Workspace --- parent it to Workspace testerPart.Position = testPosition -- give it the position end end
0
line 2 declares part as the part's position, and line 8 is cloning it. you cant clone a position dotdotdot Gey4Jesus69 2705 — 6y
0
whoops, like I said that was very quick code that I typed up just to show what I mean SteelMettle1 394 — 6y

1 answer

Log in to vote
0
Answered by 6 years ago
Edited 6 years ago

You can use something called a binary search. It's a different method of searching through tables than a standard line-by-line check, or linear search. It assumes that you have a table sorted in some manner (like a table of integers from 0-10 to provide a basic example) and then searches by cutting the results in half.

For example:

Let's say we have a table like the following

exampleTable = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10}

Now let's say we want to find the value 7. A linear search (what happens when you do "for i, v in pairs()") would go something like this. 0? no. 1? no. 2? no. 3? no. 4? no. 5? no. 6? no. 7? yes.

A binary search would go something like this:

5? no. greater. 8? no. lesser. 6? no. greater. 7? yes.

In some cases, it's not faster (for example, if the value you were looking for was 1), but in your case it might help a lot.

Some sample Lua code for a binary search can be found here

Sorry I didn't write out the code myself, I don't have the time right now. Hope this helps.

0
As soon as I read this I had to test out how many passes it would take for 10,000 records and it cuts it down to 14 passes to find the number 1, 15 pases for 20,000, 16 passes for 40,000 etc... This definitely solved my problem, thank you very much! SteelMettle1 394 — 6y
Ad

Answer this question