Creating more than 1 SubWindow MT4

hi,

Just a question: does anyone know if it is possible to create a number of SubWindows from within one indicator.

e.g. indicator: should create SubWindow 1 and a SubWindow 2, and a SubWindow 3.

I use them to plot different Objects: at the moment I create each SubWindow with a separate indicator so I though maybe anyone knows a way to create them from within one indicator.

In case it's possible and someone figured it out let me know. THANKS MJ.


pic below as example: at the moment I use 5 separate indicator to create the pic.

mt4 is not designed that way. so no.

but, if what you are trying to accomplish is having 1 indicator do a bunch of calculations and then spit them out into separate areas you can load the same indicator multiple times and then have it detect which sub-window it is and therefore what piece of data it is meant to only display.

that is to say, each indicator is capable of searching for the names of other sub-windows. using a base name and an offset index number you can uniquely self-name an indi sub-window simply by searching for a base name and iterating the index offset number until a combination is found that doesn't already exist. do this in the init section to initially give a unique name.

base = "name";
i = 0;
while (windowfind(base + i) < 0) i++;

to accommodate the likelihood that at some point a copy of the indicator will at some point get deleted (even if by accident) you need to continually re-pack the names being used. if you search for a unique unused name each time start runs, and also confirm that the index offset is less than what you previously used to self-name the indicator sub-window, you will have the indicators pack their names from 0 to copies-1 each iteration. if you don't ensure that you check that the found available name + offset doesn't use an offset less than the currently used offset what will happen is it will find it's own name in use and then find a free combination numbered higher than itself, rename itself to this higher number and therefore leave a gap in the names being used. when further copies do this same process they will fill this gap to a degree, but the last copy to run will find itself naming itself one less than it previously was, and so now the gap is the name it previously used (since the copy that originally named itself wrong would have named itself higher than the highest named offset which now JUST shifted it's own name down one offset).

now, you note the sub-window number of the current indi window using windowfind... then you iterate through the whole list of the copies of the indi you are running and compare their windowfind results.... however many windowfinds return a value less than the current copy are how many copies that currently sit above the current indi's sub-window.

i.e. your chart will be something like the following

MAIN CHART - window 0
some indi - window 1
another indi - window 2
name0 - window 3
name1 - window 4
name2 - window 5
name3 - window 6
super dooper indi - window 7

from the perspective of name2 it will eventually work-out that there are 4 copies running and the window that it happens to be running in is window 5. when it iterates through the list of indi's it discovers that name0 and name1 are in sub-windows above itself.... therefore it is the third window.... therefore it should be displaying the third type of data that you want to uniquely display within each sub-window.

similarly name0 will find no windows above, so it knows it's the first window.

in this way you have created an indicator that names itself in such a way that all names will have an offset index that are sequential, and from this you can discover where in the sub-windows the current copy sits in comparison to any other copies running.

now using a switch statement you can display the particular type of data you want based on how many copies are running above this window.

you might be asking what the reason is for having to calc how many copies are in windows above the current copy instead of just referring to the index offset number which should indicate exactly the same thing... great question! the reason you need to do this is IF you delete a copy and then load a new copy BEFORE a tick occurs and causes all windows to re-pack their names what you will have is a copy sitting below all other copies but it will automatically name itself as the number of the copy you had just deleted.

e.g.

MAIN CHART
name0
name1
name2
name3

between ticks you delete name0 and then load a new copy of the indicator the new copy will find name0 as the first available base + offset index combination and so you will have the following...

MAIN CHART
name1
name2
name3
name0

therefore, if you simply refer to the offset index to decide what the indi should be displaying you will have the data get out of whack and displayed in the wrong order. and so you must have each indi iterate through all known active copies and then find where they sit in comparison to itself. in this way name0 will still find that name1 through 3 are above itself and so it will know that it should be displaying the 4th type of data and not the first type of data once you reach your switch statement.

this is along the same lines as to why you have to continually re-pack the names each tick. if you don't re-pack the names each tick and you happen to delete more than one copy of a window and then load less than the number of copies you deleted you will again create a gap in the names being used (and so you won't know how many copies are actually running and therefore how many to iterate through when checking their sub-window positions). e.g. if you deleted 2 copies then only loaded one copy the new copy will find the free name that was used by the first deleted copy but you still have the second deleted copies name not being used which could create a gap in the sequence of names used. if you delete any copies, don't repack, and the deleted copy isn't the last copy to be loaded then again you have a gap. all-in-all you need to re-pack the names each tick to ensure any deleted copies have their names re-used by a different copy and ensure all copies are sequentially named (allowing you to iterate through all found names).

ok i've edited my last post a whole bunch of times. i will stop editing it now and leave it alone.

nubcake, thanks a lot.