Algorithm time...

[Mike Bonner]

Heres one more possibility.. If it doesn't matter which duplicate is used
if there are duplicates then the following will do 100k lines in 167
millisec.  This is using google, sorry if the blasted asterisks show up.
(To clean up yours, pasted to a field and replaced * with empty but its
still a pain in the tookus)  BTW, the search and replace in the script
editor seems to choke when trying to replace all with an empty string.

*on* mouseUp

   *lock* screen

   *put* the milliseconds into startTime

   *put* the number of lines of fld mydataNew into tcount

   *put* fld mydataNew into gbmem

   *set* the itemdelimiter to tab

   *repeat* 1000 times -- could change this to a function of the number of
lines/groupsize

      *put* line 1  to 10 of gbmem into ttemp

      *delete* line 1 to 10 of gbmem

      *sort* tTemp numeric ascending by item 2 of each

      *put* line 1 of tTemp & cr & (the last line of tTemp) & cr after tData

   *end* *repeat*

   *delete* the last char of tData

   *sort* tData numeric ascending by item 1 of each

   *filter tData without empty*

   *put* tData into field 2

   *put* the milliseconds into endTime

   *put* endTime - startTime

   *unlock* screen

*end* mouseUp


On Sun, Dec 9, 2012 at 7:59 AM, Michael Kann <mikekann at yahoo.com> wrote:

> Glen,
> In your example do you use the numbers in the left column for anything?
> Are they just the line numbers?
>
> Mike
>
>
>
> Example:
>
> 1    23
> 2    12
> 3    9
> 4    77
> 5    2
> 6    13
> 7    44
> 8    83
> 9    2
> 10  37
>
> In this example the result would be **Note if one or more values are the
> min or max the the first value found can be used
>
> 5    2
> 8    83
> _______________________________________________
> use-livecode mailing list
> use-livecode at lists.runrev.com
> Please visit this url to subscribe, unsubscribe and manage your
> subscription preferences:
> http://lists.runrev.com/mailman/listinfo/use-livecode
>