🔒 Unable to animate the change | Python | Page 1

regal oasis Oct 31, 2025, 8:34 AM

#

I am trying to make a soduko game complete in all its glory with different sizes and solver but right not the pygame part along with soduko.py are somewhat hardcoded for size=9

Main Problem:
● I was trying to animate the solver i.e display the changes it makes tho i am unable and have been stuck at this problem for well over 40 minutes.

● The soduko.py contains two classes i.e the Generator which generates the soduko itself and the Solver class which is a copy of the prior one with some essential changes i thought were necessary to animate the solving process.

Code:
https://paste.pythondiscord.com/SCAQ

cerulean duneBOT Oct 31, 2025, 8:35 AM

#

regal oasis I am trying to make a soduko game complete in all its glory with different sizes...

#

@regal oasis

Python help channel opened

Remember to:

Ask your Python question, not if you can ask or if there's an expert who can help.
Show a code sample as text (rather than a screenshot) and the error message, if you've got one.
Explain what you expect to happen and what actually happens.

:warning: Do not pip install anything that isn't related to your question, especially if asked to over DMs.

regal oasis Oct 31, 2025, 8:44 AM

#

Have i not explained the problem correctly?

rough lake Oct 31, 2025, 9:02 AM

#

Not really. You’ve basically said, “here’s a lot of code. Something is wrong.” You need to narrow down the problem and say “I think here is where something is not working correctly.”

spare heart Oct 31, 2025, 9:05 AM

#

When I wrote a sudoku creator/solver, I used Numpy arrays. 9x9x9 boolean. I ended up using tkinter, but PyGame would do.

regal oasis Oct 31, 2025, 9:09 AM

#

Problem:

● About Solver class:
-> Its supposed to solve the grid and return it everytime a change is made into it.

Thats what it was supposed to do

● Problem:
-> when solver.solve() is called in th while loop tje grid is solved in an instance which is exactly what i dont want .

-> its obvious i have to modify the solve() function but also keep it recursive so it solves correctly but i tried changing it but it got progressively worse and worse and i dont feel lile using ai to cheat so please guide me

spare heart Oct 31, 2025, 9:10 AM

#

Would writing a stack help you manage things better, vs recursive calls?

regal oasis Oct 31, 2025, 9:10 AM

#

spare heart When I wrote a sudoku creator/solver, I used Numpy arrays. 9x9x9 boolean. I ende...

Tbh i dispise the tkinter its just not very effective and tbh u can create more intricate gui using pygame in theory but at the cost of more time and writing hella alot of code

spare heart Oct 31, 2025, 9:11 AM

#

I'm not fond of tkinter, either, its primary advantage being that it's a standard library module, thus readily available.

#

Using AI is liable to get you deeper in trouble.

regal oasis Oct 31, 2025, 9:13 AM

#

spare heart Using AI is liable to get you deeper in trouble.

And even if it fixes i wouldnt know where i was wrong so i try to stay as far as possible from ai

spare heart Oct 31, 2025, 9:16 AM

#

spare heart Would writing a stack help you manage things better, vs recursive calls?

You start with a state in the stack, and every prospective step in the solve for that state gets added to the stack, testing for legality as you go. You remove the latest state, then add back to the stack each prospective step for that stack. If a state removed from the stack has no solves, the next state is removed and so on.

#

Eventually, you'll find a solution if a solution exists.

regal oasis Oct 31, 2025, 9:17 AM

#

spare heart You start with a state in the stack, and every prospective step in the solve for...

So you basically mean storing all the states upto the current one and if something goes wrong travel as back as necessary ?

spare heart Oct 31, 2025, 9:18 AM

#

Yes. The removing and discarding states with no legal moves is the backtracking.

regal oasis Oct 31, 2025, 9:18 AM

#

spare heart Eventually, you'll find a solution if a solution exists.

Well its brother is the one creating those grids

regal oasis Oct 31, 2025, 9:18 AM

#

spare heart Yes. The removing and discarding states with no legal moves is the backtracking.

Got it

regal oasis Oct 31, 2025, 9:18 AM

#

spare heart You start with a state in the stack, and every prospective step in the solve for...

Then i also need to keep track of the last change made?

spare heart Oct 31, 2025, 9:19 AM

#

Each thing added to the stack is a whole state.

#

So, fill in a number? That's a new state added.

#

Something to be built upon later when that state gets popped from the stack.

#

All you need to do is write a routine to display a state.

#

Reading the data and writing the numbers at the positions.

regal oasis Oct 31, 2025, 9:23 AM

#

So lets say the function puts a num in grid[x][y] and its wrong then it should access the last grid&change made while also removing the current one and change the num entered in the last state?
So in theory i can get away by just storing last change?

spare heart Oct 31, 2025, 9:23 AM

#

Fair warning: My solution only worked some of the time and I could never figure out quite what I was doing wrong at the time.

#

It would usually work.

#

It should have always worked, in theory, for legal starting states.

regal oasis Oct 31, 2025, 9:24 AM

#

spare heart It should have always worked, in theory, for legal starting states.

well then maybe ur generator was the one at fault?

spare heart Oct 31, 2025, 9:24 AM

#

Who knows?

#

Maybe the starting states were subtly illegal.

regal oasis Oct 31, 2025, 9:26 AM

#

hm.... for me it was checking if a num exists in a 3x3 box was hard which was causing the grid generated to be wrong hence illegal positions when removing_n digits

#

mb it was the onr for u too

spare heart Oct 31, 2025, 9:26 AM

#

regal oasis So lets say the function puts a num in grid[x][y] and its wrong then it should a...

I didn't do it by storing changes. But if you can figure out how to do it that way, okay, I suppose.

regal oasis Oct 31, 2025, 9:26 AM

#

regal oasis well then maybe ur generator was the one at fault?

well i had luck of having rndpkt helping me out

regal oasis Oct 31, 2025, 9:28 AM

#

spare heart I didn't do it by storing changes. But if you can figure out how to do it that w...

Then how?

spare heart Oct 31, 2025, 9:29 AM

#

!e py import numpy as np arr = np.ones((9, 9, 9), bool) print(arr)This was my representation of a blank grid. The first two dimensions being the 9x9 grid, the rows of Trues as the last dimension representing the remaining possible values from 0-9.

regal oasis Oct 31, 2025, 9:29 AM

#

spare heart I didn't do it by storing changes. But if you can figure out how to do it that w...

If u stored the grid with solver filling a num at [x,y] and it was wrong a fill u will have no way of finding that it was wrong unless u go back to that change and change it again go from their

cerulean duneBOT Oct 31, 2025, 9:30 AM

#

spare heart !e ```py import numpy as np arr = np.ones((9, 9, 9), bool) print(arr)```This was...

:white_check_mark: Your 3.14 eval job has completed with return code 0.

001 | [[[ True  True  True  True  True  True  True  True  True]
002 |   [ True  True  True  True  True  True  True  True  True]
003 |   [ True  True  True  True  True  True  True  True  True]
004 |   [ True  True  True  True  True  True  True  True  True]
005 |   [ True  True  True  True  True  True  True  True  True]
006 |   [ True  True  True  True  True  True  True  True  True]
007 |   [ True  True  True  True  True  True  True  True  True]
008 |   [ True  True  True  True  True  True  True  True  True]
009 |   [ True  True  True  True  True  True  True  True  True]]
010 | 
... (truncated - too many lines)

Full output: https://paste.pythondiscord.com/HLQF5MFOGKI3X3THWJSNKNKTIU

spare heart Oct 31, 2025, 9:31 AM

#

As I "confirmed" a value for that line of inquiry in a state, I'd knock out all the remaining Trues to Falses except for that one True.

#

I would then also knock out the other Trues that would imply in the same state.

#

So in that row, column, and 3x3.

#

If at any point I had no Trues left in a row or column or 3x3, that was a dead end and discarded.

#

If that makes any sense.

#

Each state was a map of remaining potential values.

regal oasis Oct 31, 2025, 9:34 AM

#

I dont mean to be judgy but the approach you took, sounds so confusing to me

First of all , how did u kept track of what num was their at place of a True or False?

spare heart Oct 31, 2025, 9:34 AM

#

As potentials were confirmed, that would knock out other possibilities.

#

True, False, False, False, False, False, False, False, False was 1 at that position.

#

But they all started off as True.

#

A blank square.

#

As numbers get filled in, the possibilities of other positions are eliminated.

regal oasis Oct 31, 2025, 9:36 AM

#

https://tenor.com/view/confused-cat-confused-math-aint-mathing-huh-cat-gif-883992111571860857

Tenor

regal oasis Oct 31, 2025, 9:37 AM

#

regal oasis https://tenor.com/view/confused-cat-confused-math-aint-mathing-huh-cat-gif-88399...

Me rn

spare heart Oct 31, 2025, 9:37 AM

#

When only one True remains, that's that square's value, assuming it doesn't get backtracked for some other reason.

#

Sorry.

#

123
456
789 Imagine a 9 x 9 replication of this.

regal oasis Oct 31, 2025, 9:38 AM

#

spare heart Sorry.

No problem man, ur approach was for you i am not that genius and its not that i am

regal oasis Oct 31, 2025, 9:38 AM

#

spare heart 123 456 789 Imagine a 9 x 9 replication of this.

Yes

#

So thats how u get the third 9 in np.ones((9,9,9), bool)?

spare heart Oct 31, 2025, 9:38 AM

#

Yes.

regal oasis Oct 31, 2025, 9:38 AM

#

Ahh.... i see what you mean now

#

So technically all the nums are prefilled in some sense then you just go round eliminating what u can, cool

spare heart Oct 31, 2025, 9:40 AM

#

As you go, if you prospect on a square having a value, you work on a copy of the state, cross out all numbers but that one, then cross out every other value that needs to get crossed out for the row, column and 3x3

#

Then if that state causes more values to be confirmed, then, still working on the copy, you cross out and potentially confirm other squares

#

and so on until it settles

#

Then you add that to the stack, because at that point you're back to prospective confirming of squares.

#

So if you wanted, you might have a set of {*range(1, 10)}

#

9x9

regal oasis Oct 31, 2025, 9:43 AM

#

Hm.... to be fair thats a very unique approach me on the other hand just relied on making an empty canvas all positions init as 0s then 1st approach was going through all cells taking a random num filling and checking if that didnt violated any rules if it did changing it and moving then i learned backtracking was needed to generate a valid grid so i factored that in

spare heart Oct 31, 2025, 9:43 AM

#

copy.deepcopy getting a workout.

#

Sometimes confirming squares will snowball

#

Confirming other squares

regal oasis Oct 31, 2025, 9:44 AM

#

spare heart Sometimes confirming squares will snowball

wdym?

spare heart Oct 31, 2025, 9:45 AM

#

spare heart Then if that state causes more values to be confirmed, then, still working on th...

.

regal oasis Oct 31, 2025, 9:45 AM

#

regal oasis So lets say the function puts a num in grid[x][y] and its wrong then it should a...

Well back to point, will the approach i shared will work?

spare heart Oct 31, 2025, 9:47 AM

#

regal oasis Well back to point, will the approach i shared will work?

Robustly implemented, I believe so.

#

But it shouldn't be random.

regal oasis Oct 31, 2025, 9:48 AM

#

spare heart Robustly implemented, I believe so.

Okay, then i am gonna go refactor the code and share the progress

regal oasis Oct 31, 2025, 9:48 AM

#

spare heart But it shouldn't be random.

Atleast not the previous num i think?

spare heart Oct 31, 2025, 9:49 AM

#

If you print out solutions as you go, you can still discard them and keep working to find other solutions.

spare heart Oct 31, 2025, 9:50 AM

#

regal oasis Atleast not the previous num i think?

Well, if you're finding solutions, first you try 1 at a position then 2 at a position and so on, accounting for legality.

#

When you backtrack...what do you try next if the thing you're backtracking didn't work?

#

The next legal thing along in a definite sequence.

regal oasis Oct 31, 2025, 9:51 AM

#

spare heart Well, if you're finding solutions, first you try 1 at a position then 2 at a pos...

Yeah thats what current solver does if its not trying to also account for returning grid for animation purposes

regal oasis Oct 31, 2025, 9:53 AM

#

spare heart When you backtrack...what do you try next if the thing you're backtracking didn'...

Well i can go back back further and further eventually reaching the start which just means instead of going back and back alltogether i should have discarded the entire solution?

spare heart Oct 31, 2025, 10:01 AM

#

regal oasis Well i can go back back further and further eventually reaching the start which ...

Reaching the start without finding any solutions would mean there were no legal solutions given that starting state.

regal oasis Oct 31, 2025, 10:03 AM

#

spare heart Reaching the start without finding any solutions would mean there were no legal ...

But the my generator code is 100% effective (i think so) so the starting state is valid ✅ then one of the nums filled between state=0 and state=current must be wrong?

#

As each box can multiple nums to be filled based on starting stat but all are eliminated except one

spare heart Oct 31, 2025, 10:04 AM

#

spare heart Reaching the start without finding any solutions would mean there were no legal ...

Or that the solving implementation was wrong.

regal oasis Oct 31, 2025, 10:04 AM

#

spare heart Or that the solving implementation was wrong.

The new one will obv have a high chance of that lol

spare heart Oct 31, 2025, 10:04 AM

#

spare heart Or that the solving implementation was wrong.

Or both.

regal oasis Oct 31, 2025, 10:05 AM

#

Is their a way to cross-check the validity of the generator?

spare heart Oct 31, 2025, 10:05 AM

#

Start with solved sudokus and knock out a bunch of numbers to be filled in.

#

That way you know there's at least one valid solution.

regal oasis Oct 31, 2025, 10:06 AM

#

spare heart Start with solved sudokus and knock out a bunch of numbers to be filled in.

I can use the backtracking part of the genertor to fill it right?

spare heart Oct 31, 2025, 10:07 AM

#

The backtracking isn't about filling so much as discarding unsolvable states.

#

I'm not sure how to make a good generator without attempting to solve it, beyond a naive legal selection of numbers at positions.

#

But not all "legal" initial states can be solved legally, maybe.

#

The starting state might not break rules in and of itself, but the numbers they force you to use, and the numbers they force you to use and so on might not result in legal states.

regal oasis Oct 31, 2025, 10:11 AM

#

Damn

#

Its just get more and more difficult/useless to make one

spare heart Oct 31, 2025, 11:02 AM

#

It's not straightforward, no.

regal oasis Oct 31, 2025, 11:11 AM

#

spare heart It's not straightforward, no.

Yeah i just started to work on it

regal oasis Oct 31, 2025, 11:28 AM

#

I dont understand this error:

Traceback (most recent call last):
  File "/storage/emulated/0/Sudoku/main.py", line 41, in <module>
    puzzle, solver_changes = solver.solve(changes=solver_changes)
    ^^^^^^^^^^^^^^^^^^^^^^
ValueError: too many values to unpack (expected 2)

@spare heart

#

Solve function

def solve(self, changes):
        empty_cell = self.find_empty()
        
        if not empty_cell:
            return self.puzzle # Means grid is full
        
        row, col = empty_cell
        
        nums = list(range(1,self.size+1))
        random.shuffle(nums)
        
        for num in nums:
            if self.check_box(row, col, num) and self.check_row(row, num) and self.check_column(col, num):
                self.puzzle[row][col] = num
                return self.puzzle, [self.puzzle, [row,col,num]]
        
        '''
        If above return statement is executed means filling is correct else we need
        to reset the last changes
        '''
        
        
        last_row, last_col, last_num = changes[len(changes)-1][1]
        self.puzzle = changes[len(changes)-1][0]
        for num in nums:
            if num == last_num:
                continue # skip to next num
            if self.check_box(last_row, last_col, num) and self.check_row(last_row, num) and self.check_column(last_col, num):
                self.puzzle[last_row][last_col] = num
                return self.puzzle, [self.puzzle, [last_row, last_col, num]]

spare heart Oct 31, 2025, 11:29 AM

#

!e py a, b, c = 1, 2, 3, 4

cerulean duneBOT Oct 31, 2025, 11:29 AM

#

spare heart !e ```py a, b, c = 1, 2, 3, 4```

:x: Your 3.14 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File [35m"/home/main.py"[0m, line [35m1[0m, in [35m<module>[0m
003 |     [1;31ma, b, c[0m = 1, 2, 3, 4
004 |     [1;31m^^^^^^^[0m
005 | [1;35mValueError[0m: [35mtoo many values to unpack (expected 3, got 4)[0m

spare heart Oct 31, 2025, 11:29 AM

#

!e py a, b, c = 1, 2, 3 print(a) print(b) print(c)

cerulean duneBOT Oct 31, 2025, 11:29 AM

#

spare heart !e ```py a, b, c = 1, 2, 3 print(a) print(b) print(c)```

:white_check_mark: Your 3.14 eval job has completed with return code 0.

001 | 1
002 | 2
003 | 3

regal oasis Oct 31, 2025, 11:30 AM

#

Fixed it

#

I had reduced the cells to be filled = 10
So it wasnt returning changes in the case all cells with 0s were filled

#

New error

Traceback (most recent call last):
  File "/storage/emulated/0/Sudoku/main.py", line 41, in <module>
    puzzle, solver_changes = solver.solve(changes=solver_changes)
                             ~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^
  File "/storage/emulated/0/Sudoku/soduko.py", line 119, in solve
    last_row, last_col, last_num = changes[len(changes)-1][1]
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: cannot unpack non-iterable int object

#

@spare heart

spare heart Oct 31, 2025, 11:33 AM

#

!e py a, b = 1

cerulean duneBOT Oct 31, 2025, 11:33 AM

#

spare heart !e ```py a, b = 1```

:x: Your 3.14 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File [35m"/home/main.py"[0m, line [35m1[0m, in [35m<module>[0m
003 |     [1;31ma, b[0m = 1
004 |     [1;31m^^^^[0m
005 | [1;35mTypeError[0m: [35mcannot unpack non-iterable int object[0m

regal oasis Oct 31, 2025, 11:38 AM

#

Hm... i tried try-except TypeError and then print(solver_changes)
It did have those 3 vals lrow, lcol, lnum

spare heart Oct 31, 2025, 11:39 AM

#

119

regal oasis Oct 31, 2025, 11:42 AM

#

Line 119:

last_row, last_col, last_num = changes[len(changes)-2][1]

spare heart Oct 31, 2025, 11:43 AM

#

You're running this in on your phone? Well done. I hope you're not writing this on your phone. That would suuuuck.

regal oasis Oct 31, 2025, 11:43 AM

#

Yes, i am running this on my phone

spare heart Oct 31, 2025, 11:44 AM

#

regal oasis New error ```py Traceback (most recent call last): File "/storage/emulated/0/S...

changes[len(changes)-1][1] has a type of int

regal oasis Oct 31, 2025, 11:45 AM

#

Soo for some reasons the errors are popping occasionally

#

It solved this

spare heart Oct 31, 2025, 11:46 AM

#

If it helps, indices may be negative.

regal oasis Oct 31, 2025, 11:46 AM

#

spare heart `changes[len(changes)-1][1]` has a `type` of `int`

But it does contain those 3 values to be unpacked , then why error?

spare heart Oct 31, 2025, 11:46 AM

#

The error message is telling you that it doesn't.

regal oasis Oct 31, 2025, 11:46 AM

#

spare heart If it helps, indices may be negative.

Yeah thats correct actually didnt thought of it

regal oasis Oct 31, 2025, 11:47 AM

#

spare heart The error message is telling you that it doesn't.

But i print it and then puff it has

spare heart Oct 31, 2025, 11:47 AM

#

Are you printing that, or something else?

regal oasis Oct 31, 2025, 11:47 AM

#

And also its not getting TypeError anymore instead its ValueError

regal oasis Oct 31, 2025, 11:47 AM

#

spare heart Are you printing that, or something else?

Printing

#

try:
        puzzle, solver_changes = solver.solve(changes=solver_changes)
    except ValueError:
        print(solver_changes[1])
        break

spare heart Oct 31, 2025, 11:48 AM

#

No, as in you are printing, that or something else.

#

Damn it...

#

um....

#

You are printing either that or printing something else.

#

Be sure you know which.

regal oasis Oct 31, 2025, 11:50 AM

#

Traceback (most recent call last):
  File "/storage/emulated/0/Sudoku/main.py", line 43, in <module>
    sudoku_surface = GenSurface(surface_x,surface_y,size).make_surface(puzzle)
                              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/storage/emulated/0/Sudoku/soduko.py", line 119, in solve
    last_row, last_col, last_num = changes[len(changes)-2][1]
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ValueError: too many values to unpack (expected 3)

regal oasis Oct 31, 2025, 11:50 AM

#

regal oasis ```py Traceback (most recent call last): File "/storage/emulated/0/Sudoku/main...

I removed the try-except

#

See its line 119 and the same index for change

#

Oh..fcm

#

Fck

spare heart Oct 31, 2025, 11:50 AM

#

Mhm...

regal oasis Oct 31, 2025, 11:50 AM

#

I am printibg wrong

#

Okay somehow its indexing the puzzle and not the rows,col,nums part

#

pithink

#

[1, 5, 9, 7, 3, 8, 4, 2, 6]

#

Oh.... see this

[4, 7, 3, 6, 1, 5, 2, 9, 8]
[[5, 2, 8, 3, 4, 9, 7, 1, 6], [4, 7, 3, 6, 1, 5, 2, 9, 8], [6, 1, 9, 2, 0, 8, 0, 5, 4], [0, 4, 0, 5, 7, 3, 0, 0, 0], [3, 0, 7, 0, 6, 2, 4, 8, 1], [9, 0, 6, 1, 8, 4, 5, 3, 0], [2, 6, 0, 4, 0, 1, 8, 7, 0], [0, 0, 0, 8, 5, 0, 1, 0, 9], [8, 9, 1, 0, 0, 7, 0, 4, 5]]

it doesnt have the second list containing last change

#

@opalmist

spare heart Oct 31, 2025, 11:56 AM

#

I'm here. I just didn't have anything to say.

regal oasis Oct 31, 2025, 11:57 AM

#

Same

regal oasis Oct 31, 2025, 12:22 PM

#

Yessss

#

Its working

#

See it for ur self
https://paste.pythondiscord.com/3OJQ

#

@spare heart

spare heart Oct 31, 2025, 12:24 PM

#

Neato burrito.

regal oasis Oct 31, 2025, 12:29 PM

#

You like it?

#

Now one last thing i guess

#

Thank you @spare heart

spare heart Oct 31, 2025, 12:31 PM

#

regal oasis You like it?

I am on my phone.

#

I think if you've made it work reliably, good.

#

Try to trip it up.

regal oasis Oct 31, 2025, 1:33 PM

#

spare heart Try to trip it up.

?

#

Find errors u mean?

#

But look at its time complexity

regal oasis Oct 31, 2025, 1:54 PM

#

Lol my implementation beat someone elses on internet and the other one is still solving it py_strong

#

#

@spare heart

cerulean duneBOT Oct 31, 2025, 3:01 PM

#

Python help channel closed for inactivity

This help channel has been closed. Feel free to create a new post in #1035199133436354600. To maximize your chances of getting a response, check out this guide on asking good questions.

#🔒 Unable to animate the change