Thread: How to shift elements of an array?

Stacks are FILO structs in my way of thinking - Queue's are FIFO, and I don't see a really striking resemblance to either one, here.

One more vote for the for or while loop. Any elegance it may appear to lack compared to something like a circular buffer, is more than made up for by the supreme elegance of it's intuitive simplicity.

Symantec arguments aside, how on earth is the case where there is a 4096 value container and two pointer re-assignments and two value assignments somehow less elegant than 4094 reads and 4096 writes? wow.

Originally Posted by jeffcobb

Symantec arguments aside, how on earth is the case where there is a 4096 value container and two pointer re-assignments and two value assignments somehow less elegant than 4094 reads and 4096 writes? wow.

How about every other case where I actually want to do anything else with the data? Like for example, finding one specific element?

Code:

array[ X ] = 5;

You know, every other normal operation you'd be doing...


Quzah.

Originally Posted by Subsonics

But since you can do operations on both side of it like suggested, for example pushbase popbase and the ordinary push and pop. It would not follow the LIFO principle, at least it should be signified in some way as a special case of a stack, no?

You're describing a "Deque" (Double-ended queue).
These can either be used as a stack or a queue, or both at once (useful for some algorithms I know of). Calling it "a stack" might be wrong, but only in so far as it's much more than that.
Point taken anyway, ta.

Originally Posted by Subsonics

But since you can do operations on both side of it like suggested, for example pushbase popbase and the ordinary push and pop. It would not follow the LIFO principle, at least it should be signified in some way as a special case of a stack, no?

Originally Posted by Adak

Stacks are FILO structs in my way of thinking - Queue's are FIFO, and I don't see a really striking resemblance to either one, here.

Now, I know y'all ain't that dumb. Perhaps I should have said "stackish" and "que-esque". No one is going to write something and go, "Well, I can't add a pop from bottom because that disobeys the laws of FILO" or something.

Originally Posted by quzah

How about every other case where I actually want to do anything else with the data? Like for example, finding one specific element?

Code:

array[ X ] = 5;

As mentioned several times already, X would be retrieved via a get/set() to do the dead simple arithmetic from the current first element. If you want the 5 element, that's first_element+5-1.

Haven't you people done any actual programming? Or are you recommending one intentionally stick one's head in the sand whilst coding? Sheesh.

Originally Posted by MK27

As mentioned several times already, X would be retrieved via a get/set() to do the dead simple arithmetic from the current first element. If you want the 5 element, that's first_element+5-1.

Haven't you people done any actual programming? Or are you recommending one intentionally stick one's head in the sand whilst coding? Sheesh.

BFD. I don't care if you use get or set, it's still slower than me using an array. Which was the whole point of my post. You using a function call to access some member, no matter how you do it, isn't ever going to be faster than my array access. EVER.


Quzah.

Well since the OP specified operating on the ends of the container and once you have more than a few items, doing two operations will always be faster than two + any extra operations to rotate the contents. Or as you put it ALWAYS. Heh. Since no operations on the middle of the container were specified, none were considered and any extra operations on the middle for the simple purpose of using a specific type are by definition wasted.

array[ x ] = 5;

Is always going to be faster than:

put( 5, x );

It doesn't matter if it's at the beginning, end, or middle. None of that actually matters though, because none of the implementations suggested, by anyone, actually do what they asked to do:

Shift an array, and zero fill the items that were shifted away.

Furthermore, the OP hasn't responded at all to this thread, so the speculation as to "only the ends are accessed" isn't accurate at all. They didn't say anything about only accessing end members.


Quzah.

Actually we all screwed the pooch on this one, including you Quzah. Before you hit the reply button in indignation, reread what the OP wrote:

How would I then use the value to produce a second array that has the depth from 1 to 3cm and then have zeros in the elements after 3cm so that the new array is the same size as he original?

IOW he wanted a second array, not a shifted version of the first one. Stacks and for loops and array access suddenly mean little which is probably why he/she never responded. We were giving him answers to a question he did not ask. Heh.