How to find prices closest together?

I'm trying to find a way to do this. At first I tried to do something using MathAbs and substracting values from each other, but it seemed I was oversimplifying it since I'm dealing with like 20 prices here. I already have them declared and all, so it's just a matter of finding the correct ones.

Is there a quick and easy way to take such a number of prices and get the price difference of every possible combination (or find some way to only get the combinations that matter)? Or perhaps filter them by some arbitrary maximum value (that if the difference exceeds it the price is ignored)?

One way to do it:
Make an array and fill it with all prices you want to check.
Sort the array, this way you get them in nice ascending or descending order
with 1 command.

declare a double LowestDistance = 9999999.99;

start at one end of the array and compare the prices like this:

if(price1-price2 <= LowestDistance) {LowestDistance=price1-price2;
price_a = price1; price_b = price2; }
if(price2 - price3 <= LowestDistance) {LowestDistance = price2-price3;
price_a = price2; price_b = price3 }
if(price3 - price4 <= LowestDistance) {LowestDistance = price3-price4;
price_a = price3; price_b = price4 }
.....
....

This can be done in a loop.

After the end of the loop, you will have 3 fields filled with the correct
values:
LowestDistance
price_a
price_b

I think that is what you need.

Thanks. So I make an array and then a loop that loops through the arrays index and compares the values in the indices?

Do you have any idea how I would have to adjust in the case that there's 4-5 values relatively close together (say difference, but it could be any amount below a specified maximum: 0.00005, 0.00004, 0.00003, 0.00002). It's not that I specifically need the 2 prices closest together, but a number of prices that fits the maximum difference?

Would this be a good solution to that issue?
If I am using arrays correctly like this we should end up with a 2 dimensional array, and the first dimension contains 20 prices and the second either 20 or 0 to 19 integer. If the integer is 20 no price is close enough, if the price is 0-19 then this corresponds with the index of the price that is close enough to this price. Although the index can be overwritten by the next price the other price will still contain the index and I could possibly add in a check that compares the difference of price 1 and price 3 (price 2 being within maxdiff of both price 1 and price 3) before setting the second dimension of the array (if (Array == 20), something like that).

Pipsgeek, this is different from what you asked.

So you want to define a maximum distance and find all price-pairs
which are that close to each other ?

There is more than one way to do it.

One would be:
Make a 1 dimensional array and fill it with all prices you have.
Sort it ( ascending or descending, doesn´t matter ).

Loop once through the array and check if distance between those 2 prices <= maximum_difference

If yes, do something with that value.
Either store it in another array ( or do the whole check in a 2 -dimensional array where index 0 contains price and index 1 contains
the distance to the previous price )

From this check you have all price levels extracted which are closer to each other
than your defined maximum_difference (or equal to max diff ).

I cannot say more because I have no idea why you are doing this
and what you will do with the extracted price-levels.
But that is necessary to know how to store the levels and what additional info is needed to work with these.

The example I posted won't work?

Perhaps to make it clearer, the prices are possible targets of future prices based on various indicators. I want to find out when multiple of these possible targets are close together and base entry/exits on these prices. How I'm going to define "close together" is an entirely different issue that is likely to be able to be solved using ATR or something similar, or just an external value that I can adjust as volatility changes.

I realize it isn't the same as what is asked in the thread title, but the problem is I'm not 100% sure how to go about it yet. Halfway through I realized it would be better to define a maximum because it could be possible that more then 2 prices are close together.

I'm interested in how to basicly have the setup and that I can go a different way from there. If the code I posted works as I think it works then this should be enough for what I want to do. I don't have any experience with Arrays other the series used in indicators so I'm not exactly sure what the correct way for working with the array is. I figure you just go like Array [Dimension1] [Dimension2] and you have the value that is at those coördinates. I just omitted the repetitive declaration of all the prices in the array (Price0, Price1 should go on till Price19).

For instance, is double Array[19][1]; correct or do I need to use [20][2] because zero isn't counted when declaring it? (it should create an array that is 20x2, and if I'm assuming correctly the index should start at 0 and end at 19 and the second dimension indices should be 0 and 1)

I haven´t used arrays for about 1 year and forgot how to use them.
In MetaEditor there is a Help button and from there you can find how to declare arrays etc.
Or here: http://docs.mql4.com/

Sorry, I don´t have the time to do the reading for you

lol, thanks for giving me the idea to use arrays. I figured arrays were peanuts for you and you could immediatly see if it was correct or incorrect. From the rather unclear MQL4 help examples it seems the correct way to declare it would be to use the number of values, not the index numbers (so Array[20][2] for a 20x2 array).