Scratch

Please do. And Cool!

Thank you! This project was far more successful than I'd have expected! Thanks everyone!

bharvey wrote:

coolstuff wrote:

Then, by starting on the starting point, you need only go towards the smallest number, because smaller the number, the closer you are to the goal. In this case, you would take 19, 18, 17, 16, 13, 8, 3, then the goal.

I can see why taking the smallest neighboring number gets you closer to the goal, but it's not obvious to me that taking the smallest number right here always gives you the shortest path overall.  (Since you give different unique numbers to squares equally far from the goal, maybe the overall shortest path isn't the one that you found first on your way outward from the goal.) Can you prove this is optimal?

I can't prove it, but you'll notice from the project that it works. If you think about it, it works quite well - because you're finding the fastest way from the "you" square to the goal, it works perfectly. It's hard to explain.

And I'm sorry, but I know only how to make Scratch projects work faster.

CHEZA69 wrote:

This is the work of a genius really awesome project

Thanks! That really means a lot to me!

This is a REVOLUTIONARY project! Everyone, thou shalt see it! Discover what everyone's raving about:

Keleymeister: "Wow, that works really well!"

Addzero: "This works well, nice work!" "Neat project" "Awesome job!" "Very cool!"

AnnoyingOrange: "This is a fun game! Great job at making everything!"

Paddle2See: "What a wonderfully elegant algorithm and project!  Very impressive " "Wow...that's really impressive!" "Lot's of fun! The only way I could confound it was to give it an impossible situation  "

CHEZA69: "This is the work of a genius really awesome project" wow thats really impressive with an impressive script to go with it

Lucario621: "very impressive! It works great!"

imacat: "woah."

Jonathanpb: "That's brilliant, coolstuff!"

Shadow_7283: ":0... This is probably the best pathfinding AI on Scratch. I've made several attempts at a similar AI, but they have never worked even close to as well as this."

RHY3756547: "Wow - this beats lucario's by like, a thousand miles. Great job! Never got stuck in any of the thousands of dead ends I set up.  "

iCode-747: "Amazing!"

Illusionist: "Amazing!"

aweman: "INCREDIBLE! I made a simple maze for it, and it did it INSTANTLY!"

Kewlbeanz: "nice work  "

Frydad: "Geez, this is cooler EVERY SINGLE TIME. I'm just plain impressed…"

Jimothy: "awesome! this gives me a lot of good ideas!"

My brother, Mathwizz, Made one that is praticly flawless. This is the link http://scratch.mit.edu/projects/MathWizz/803671

I've updated it so that the loading is even faster than before. I also fixed a few bugs, but also made it so that the 1-sprite-1-scriptness was gone. The script just got too long to work with!

Any suggestions for future updates would be greatly appreciated!

very cool! what algorithm did you use? A*?

well i'm trying to open in scratch, but the script is so large i can hardly do anything cause scratch is lagging a lot.

that'd be great if you could, not sure how much i can help, but i can try

Hm... I tried taking a picture of it, but Scratch freaked out at me - sorry!

Bumpetyboo.

<say[omg im on scratch fourums( <cool< <=>true  )