Finding maximum value of dynamic parameter

Hi,

Could do with some advice here...

My Aim/Motivation:
I have a parameter which is constantly updated throughout the code. I want to find the value for this parameter at its maximum.


I want this to be as fast as possible and I dont want to store much data if I can help it.

I am sure things like this have been done many a time before. So I am seeking the expert advice of this forum.

Any Ideas/Suggestions/websites/Examples? most welcome!

thanks!!!

The most straightforward solution to me involves creating a class for which this variable is a member, and the largest encountered value of this variable is another member. So, you control the updating of this variable such that when a larger value is assigned to it, the other member is updated.

You need to provide some concrete example of what you mean. Is this "parameter" a global variable? A member of a class? I recommend making it a private member of a class, and have the set function test to see if is maximum and store the value in another private class variable if it is the maximum. By encapsulating this variable in a class you can enforce the checking for the maximum value whenever that variable is changed.

Edit: too slow.

Jim

He could mean parameter as in an argument to a function, in which case I think he's looking to find the maximum in the wrong place. It would be better to find out where the parameter came from and do some analysis, rather than try to do it from inside the function. How good your answer is depends on how many times the function was called, otherwise.

Thanks to both!

This parameter is relating to maximum position of a body (from a given reference) which is constantly being updated due to other equations being iterated which determine the bodies motion.

So this parameter is constantly being updated within a while loop.

Can you offer an alternative suggestion to using a class?

thanks again

Quote:

---

Originally Posted by strokebow

Can you offer an alternative suggestion to using a class?

---

What do you find deficient about the suggestion that you would like an alternative?

EDIT:

Quote:

---

Originally Posted by strokebow

So this parameter is constantly being updated within a while loop.

---

Oh, then another possible solution is to keep track of the largest known value in the code surrounding the while loop, i.e., "constantly updated throughout the code" is not actually the case.

Well laserlight's idea still works, but instead of writing a class you would declare another variable for the maximal value and then call std::max() over and over in the loop. When the loop is done, just look at that maximal value.

Quote:

---

Originally Posted by whiteflags

Well laserlight's idea still works, but instead of writing a class you would declare another variable for the maximal value and then call std::max() over and over in the loop. When the loop is done, just look at that maximal value.

---

Thanks!!

So basically, just keep a variable for the max.

I suppose its quicker to just do:

Code:

---

                        xMax = (X > xMax)?X:xMax;
                                yMax = (Y > yMax)?Y:yMax;

---

Whereby, if X is greater than xMax then the value for X becomes the new value for xMax otherwise xMax keeps its current value. And similarly with Y?

What do you think? OK?

FYI, there is a function named std::max you can use if you want. Does the exact same thing.

Furthermore,

How would I do this in the fastest possible way to also take account of negative values.

i.e. the maximum value irrespective of sign.

Would it be just a case of:

Code:

---

xMax = (abs(X) > xMax)?abs(X):xMax;
        yMax = (abs(Y) > yMax)?abs(Y):yMax;

---

Or is there a faster way of doing this?

thanks again

Quote:

---

Originally Posted by strokebow

How would I do this in the fastest possible way to also take account of negative values.

---

I don't know about fastest, but simplest:

Code:

---

xMax = std::max(xMax, std::abs(X));

---