Can anyone tell me why the 2 scripts that i attached are giving different outputs for the same inputs?

Can anyone tell me why the 2 scripts that i attached are giving different outputs for the same inputs?

They should be doing exactly the same and it's crushing my brains.

The second script also seems to be slower and it's graph is less good than the first script, I have no clue why.

[The Legend]

bump

[John D'Errico]

Not just two scripts. Two families of scripts, burried 4 levels deep in folders, that have no documentation at all, so we need to guess what it is they should be doing. Sorry, there is a good reason why you've gotten no response so far.

If you want help, then make it easy for somone to help you.

[Adam Danz]

Hi, The Legend , I un-flagged John's comment.

His advice will help you if you follow it.

The idea in these types of help-forums is for the person asking the question to do 95% of the work so that we can very easily and very quickly see the problem and spend our time investigating the problem instead of spending loads of time trying to figure out where the probem is or how to reproduce it. Remember that we are unpaid volunteers.

That means that you might have to restrcuture your code or data in order to frame it as a minimal working example even if that takes you a lot of time to do. An idea situation would requires us:

• 30 seconds to copy/load code & data into our systems
• 30 second to run the code and reproduce the problem
• a few minutes to fully understand the problem and come up with a solution (though, usually more)
• a few minutes to type it out (though, usually more)

[Meg Noah]

You're initializing the boards with random numbers, but did not use a seed to ensure that both scripts operate on the same data, it seems.

[John D'Errico]

I'm not sure what I said that bothered the legend. But the point is, it can easily take far more time than just a few minutes to figure out what is being done, then digging deeply into the code, finally explaining what the problem is. When you make it difficult for someone to figure out what you are doing, that makes it just that less likely that someone will bother to make the effort to tackle your problem.

Make it easy for somone to help you. That means you need to tell them what you are trying to do, and what result you expect. In a case like this, you need to explain why you see a difference, and why you think it is significant. Is the difference just a difference in the least significant bits of a number (often the case) or is it something big?

One thing you should consider is that completely undocumented code is just a bunch of apparently random commands, doing whatever you decided to do. Figuring out what the code should do from just the code can be very difficult, especially when there are many functions involved calling each other.

Remember, you are asking someone (who is wholly unpaid to do this task) to invest potentially as much as an hour or more of their time on you. I can often spend more than an hour or more writing a response to a question, especially when followups are considered. By making it easy for someone to help you, you make it more likely that someone will invest that time in you.

https://www.mathworks.com/matlabcentral/answers/6200-tutorial-how-to-ask-a-question-on-answers-and-get-a-fast-answer

[Meg Noah]

I don't know if this will fix it, but when you run monte carlo operations twice, without reinitializing the random number stream seed, then the algorithms will be operating on two different data sets and necessarily will produce different results. Depending on where and how the random numbers are called, you want to be sure they are solving the same set of mazes. If the only place rand is called is in the generation of a new maze,then executing this once in the main program should result in getting the same set of mazes to solve - you need to check it. The monte carlo solution should have a seed value that is the same. Here, it is '1' but you can make it any number even 8675301 if you want.

s = RandStream('mt19937ar','Seed',1);
RandStream.setGlobalStream(s);

If there are other places where rand is called, then it's going to be more complicated because you want the two algorithms to solve the same maze. Ideally, have the maze generated in a main,, and then send it to the different solvers all in one program.