Re: Sort problem
- Posted by don cole <doncole at pacbell.net> Jun 12, 2007
- 548 views
CChris wrote: > > I'd expect some sort of corruption occurred in your actual dataset, at least > in the groups with unexpected results. More precisely, I suspect some records > don't have the same length. > > Could you please check the actual individual entry with movie number "D1000" > in the solddvd group, and see if it has the same length as the one with > number "D0999"? I expect it to be shorter. At this point, there may be extra > data in the "last" item or missing data in the "first" one just as well. > Or wrong labelling, causing a dvd to be in a nondvd category or vice versa. > > CChris Thanks to CChris and Pete, I think you are both right about my data being corrupt. That's what got me stared on this in the first place. What I didn't show you was my complete window which is: number first last movnum db letter history of movie movie length length length movies number number errors errors errors errors ------ ------ ------ ------ ------ ------ ------ reg 271 3119 4907 0 0 0 0 adult 24 A4730 A5582 0 0 0 0 dvd 330 D1031 D4234 0 0 0 0 soldreg 2514 0888 5150 0 0 0 0 soldadult 2527 A1121 A0985 0 0 0 0 solddvd 3164 D1000 D0999 0 0 0 0 Let me go back and check my error checking and try Pete's experiment. Don Cole