Reuse fractal formula iterations

[flatFeather]

Reuse fractal formula iterations. Issue #30

[MuhammadMuradG]

Hi, @flatFeather! This PR seems to be a duplication of #61, is this correct or there is a difference?

[flatFeather]

Hi @MuhammadMuradG, It is correct. Unfortunately it seems that someone and I have been working on the same issue. I think I told you that I was going to be with this issue. It is very difficult for me to continually check who is with what, we should find a way to centralize that.

[flatFeather]

Same approach taked:

• Implemented a caching mechanism (self.computed_values) to store intermediate results of points computed during fractal generation. The caching mechanism checks if a point has been computed. If so, it retrieves the cached value instead of recomputing it.
• A typo in gui.py has also been corrected: Este texto tiene una palabra formated_fractal formatted_fractal

[flatFeather]

After some manual performance tests: It takes 15 seconds to go from one derivation level to another, generating the images. After this code it takes only 5 seconds.

[MuhammadMuradG]

Hi @flatFeather!

You can simply if you intend to work on specific issue create a PR that describe what you intend to do before even create any valuable commits then convert it to draft which is mean the work in progress... However, the contributor of PR 60 did not address all the issue raised and it seems that he will not continue work on his PR.

Also, you wasn't need to create multi consecutive comments to describe your PR instead you can edit PR description box.

[MuhammadMuradG]

It seems you are on the right way, I will left some point you need to address inline.

[MuhammadMuradG]

I gave it a new shot of review, and my previous suggestion about "resetting the attribute" will introduce a bug relating to the fact that this suggestion discarding the different between p_params being used and then I realized that we need to move z = c code line before if statement but after making this change I can not find any performance enhancement, so, how you tested the enhancement performance?

[flatFeather]

Here how I tested the enhancement performance:

https://github.com/Yuri-SVB/Great_Wall/assets/50509521/25a8d87e-874f-423b-9da9-51d642f779d2

[Yuri-SVB]

Muhammad seems to be right: the last change seemed to change nothing.

[flatFeather]

If the value of _pparam is updated in the update() method during the app flow and you want that change to be reflected in the fractal calculation, then you will need to recalculate all the points in the set each time _pparam is updated. This means that you cannot use the cache to store the previous results, since you need to recalculate the values with the new _pparam.

[flatFeather]

I have modified the code so that when using the cache it takes into account the _pparam to be used or not. But the use case is very corner case, I could not even reproduce this case to see the time difference. There is no difference in the normal flow of the app.

[MuhammadMuradG]

Since this PR doesn't introduce any enhancements, is there still a need to merge it? However, I think you may have been able to conclude that this approach wouldn't yield significant improvements if you had think in the potential benefits beforehand.