Re: Array size limit ?

Tech-Archive recommends: Repair Windows Errors & Optimize Windows Performance


"DanS" <t.h.i.s.n.t.h.a.t@xxxxxxxxxxxxxxxxxxxxx> wrote in message news:Xns9AC56AFF243AFthisnthatadelphianet@xxxxxxxxxxxxxxxxx

That's all fine and dandy, but just setting up the arrays is barely
anything compared to the real number crunching after they are set up.

Like I described in my OP, we're working on a table with 65,535 columns
and 162 rows. Each column is a hop pattern and contains numbers between 1
and 128, obviously duplicated since there are 162 rows and only 128
numbers. So col1/row1 is the first channel of hop pattern 1, col1/row2 is
the second channel of pattern 1, etc.

1 45 34 27 ....(65,535 columns)
34 76 43 23
21 82 82 47
26 71 101 89
.
.
.
(128 Rows)

The excercise is to identify hopping patterns (columns) that are unique
to each other in that they never fall on the same channel in the same hop
(row). In the example above, Pattern2(Col2) & Pattern3(Col3) both hop on
channel 82 on the third hop, so those 2 are eliminated.

There's 2 way to do this, one thru a brute force method, comparing each
of the 65K cols to all of the others on a one-to-one basis,


Well if you consider the first row, there can only be 128 unique columns at most, in fact there can only be 127 as there has to be at least 65,408 columns that have the same hop value. So clearly comparing the first row is going to be the big eliminator. In which case I'd probably put those values in a single array (0 to 65534), do a quicksort on an indexer based on those keys, then step through eliminating duplicates. That'd be Ologn + n.
Again there is still no major number crunching, just simple retrieval, and comparisons.

or some type
of 'genetic' search, of which I don't have the details.


I doubt that's needed. This isn't a difficult of timely task.


As I said, I'll have more details after I see code.


Look forward to seeing it.








.

Relevant Pages

  • Re: Array size limit ?
    ... So col1/row1 is the first channel of hop pattern 1, ... Well if you consider the first row, there can only be 128 unique columns at most, in fact there can only be 127 as there has to be at least 65,408 columns that have the same hop value. ... So clearly comparing the first row is going to be the big eliminator. ... Again there is still no major number crunching, just simple retrieval, and comparisons. ...
    (microsoft.public.vb.general.discussion)
  • Re: A "slanted edge" analysis program
    ... sense to compare the edge *position* of each channel with the edge position in the luminance channel? ... comparison of a color channel with Luminance. ... Would comparing the three channels to luminance, instead, have any advantage over SFRWin's approach? ...
    (comp.periphs.scanners)
  • Re: Distortion on right channel
    ... with any set of speakers; ... replacing electronic components in circuit boards (when I say me I ... it's the same as the other channel, make a mental note, and move on to ... As a beginner you'll need to start brute-force, comparing and often ...
    (rec.audio.tech)
  • Re: Psystar Macs Better Than Apple Macs?
    ... Motu or comparable) is superior to a two channel 24 bit 192 kHz unit ... I'm glad I can hear sound and it is pretty good on the Apple. ... how much the sound card costs -as long as we are comparing oranges to ...
    (comp.sys.mac.advocacy)
  • Re: A "slanted edge" analysis program
    ... (http://www.imatest.com/docs/sfr_instructions2.html almost halfway down the page under Channel). ... And Luminance weighting according to ITU-R BT.709 is: ... Would comparing the three channels to luminance, instead, have any ...
    (comp.periphs.scanners)