How to get 5x5 scramble from a diagram?

[abunickabhi]

The 5x5 scramble is unknown, but I do have the 5x5 cubestate that I can paint, by seeing a video.
How do I get a scramble (a 20-30 move sequence that gets me to that state from solved state) if I just specify all the stickers on the 6 sides?

(I do such a tool exists for 3x3, Hagis : Here's a Graphic Interface to the Solver or just plain Cube explorer does the job for the 3x3 puzzle.) I need a tool for higher NxN puzzles.

 

[Hazel]

This requires the computer to solve the scramble very efficiently, then to reverse the scramble. This can't be easily done I don't think, which is why the WCA uses random moves for 5x5 and up instead of random-state. I'm sure there's solvers out there that can solve 5x5s somewhat efficiently, but I'd be surprised if you could find one that can consistently do it in under 30 moves :/

 

[qwr]

What is the most move efficient big cube solver right now?

 

[abunickabhi]

What is the most move efficient big cube solver right now?

I think the best 5x5 solver out there is https://cubesolvingprograms.freeforums.net/thread/11/omnia-obtorquebantur-5x5x5-notes-updates. Its made by Ed Trice.

@xyzzy can confirm maybe.

 

[xyzzy]

What is the most move efficient big cube solver right now?

Five-phase redux; I wrote the first implementation, Chen Shuang wrote his own independently (and it's better than mine). dwalton76 has a simpler but slightly less efficient variant implemented in his solver.

I think Ed's solver might be better if you're using the same kind of computers he's using (literally hundreds of gigabytes of RAM and dozens of CPU cores). It's not exactly something you just go and use on a normal desktop.

---

Ideally you'd want some kind of graphical program where you can paint in the facelets (like a big cube version of Cube Explorer), but unfortunately this doesn't exist yet. It might be possible to rig dwalton76's solver to do this (it can already recognise facelet colours from photos). If you don't want a graphical interface, any of the solvers mentioned above will work just fine.

 

[qwr]

sounds like a fun coding problem. if I weren't so lazy I'd try to make an efficient C++ solution. But I would make one big program and have 5x5 as one case, not a separate program for 5x5

 

[abunickabhi]

Five-phase redux; I wrote the first implementation, Chen Shuang wrote his own independently (and it's better than mine). dwalton76 has a simpler but slightly less efficient variant implemented in his solver.

I think Ed's solver might be better if you're using the same kind of computers he's using (literally hundreds of gigabytes of RAM and dozens of CPU cores). It's not exactly something you just go and use on a normal desktop.

---

Ideally you'd want some kind of graphical program where you can paint in the facelets (like a big cube version of Cube Explorer), but unfortunately this doesn't exist yet. It might be possible to rig dwalton76's solver to do this (it can already recognise facelet colours from photos). If you don't want a graphical interface, any of the solvers mentioned above will work just fine.

dwalton's solver works perfectly for me.
https://github.com/dwalton76/rubiks-cube-NxNxN-solver

I just solve the cube in the diagram, and reverse it to get the scramble.
Life is much easier now.