fastest data structue

[Loui Mercieca]

Hi,

In my design i have a data structure used to store large amount of
numbers ( in the range of lots of thousands ). Each element contains 3 items
and the no of elements are dynamic.. of the form:
[X1,Y1,Z1]
[X2,Y2,Z2]

 

[Joep]

You should consider how you want to access the elements first. Is there any
order or relation between the elements?

 

[Loui Mercieca]

First In First Out queue..

There is no need for sorting or searching or any data manipulation.

You should consider how you want to access the elements first. Is there
any order or relation between the elements?

Hi,

In my design i have a data structure used to store large amount of
numbers ( in the range of lots of thousands ). Each element contains 3
items and the no of elements are dynamic.. of the form:
[X1,Y1,Z1]
[X2,Y2,Z2]
.
.
.
[Xn,Yn,Zn]

The topmost priority is time, but memory efficiency is important too..
Can anyone the most suitable way of implementation.

Thanks

 

[Joep]

A queue is what comes up first since that is a fifo collection. Maybe
storing structures instead of objects increases speed a bit. What will you
be doing with the 3D points?

First In First Out queue..

There is no need for sorting or searching or any data manipulation.

You should consider how you want to access the elements first. Is there
any order or relation between the elements?

Hi,

In my design i have a data structure used to store large amount of
numbers ( in the range of lots of thousands ). Each element contains 3
items and the no of elements are dynamic.. of the form:
[X1,Y1,Z1]
[X2,Y2,Z2]
.
.
.
[Xn,Yn,Zn]

The topmost priority is time, but memory efficiency is important
too.. Can anyone the most suitable way of implementation.

Thanks

 

[Jon Skeet [C# MVP]]

A queue is what comes up first since that is a fifo collection. Maybe
storing structures instead of objects increases speed a bit.

On the contrary - that would *decrease* performance in the queue (until
2.0 with generics) as it would require boxing and unboxing. Even with
generics, using objects is likely to be faster for putting things into
the queue and removing them, as a reference is smaller than three
integers (or doubles, or whatever).

 

[Loui Mercieca]

I created a dll, which accepts data from a port ( serial/ parallel/ usb )
and fills this 'array', thus i must ensure that every point is handled and
stored. What i really don't want is that some data on the port is lost due
to storing latency.

However the area to be scanned is quite large ( 500mm/500mm/100mm at a
resolution of 0.02mm ) thus i would like to use memory efficiently

I use this 'array' then to plot these points as a cloud mesh, at a later
time..

A queue is what comes up first since that is a fifo collection. Maybe
storing structures instead of objects increases speed a bit. What will you
be doing with the 3D points?

First In First Out queue..

There is no need for sorting or searching or any data manipulation.

You should consider how you want to access the elements first. Is there
any order or relation between the elements?

Hi,

In my design i have a data structure used to store large amount of
numbers ( in the range of lots of thousands ). Each element contains 3
items and the no of elements are dynamic.. of the form:
[X1,Y1,Z1]
[X2,Y2,Z2]
.
.
.
[Xn,Yn,Zn]

The topmost priority is time, but memory efficiency is important
too.. Can anyone the most suitable way of implementation.

Thanks