extrapolating data

[silky_green]

let me first apologize for my ignorance of both excel and statistics.
have data that i'm trying to extrapolate. let me just share this dat
with you:

XXX YYY ZZZ
100 649 46439
200 1531 61384
500 5485 89121
1000 4117 129452

what i'm trying to plot is what the values of YYY and ZZZ will be i
XXX grows, say by increments of 500:

XXX YYY ZZZ
100 649 46439
200 1531 61384
500 5485 89121
1000 4117 129452
1500 ???? ???????
2000 ???? ???????
etc.

or i guess another way to look at this is: what are the values of XX
and YYY when ZZZ reaches 250000?

thank you

 

[AlfD]

Hi!

A bit thin on data...

Are you able to plot charts of XXX against YYY and XXX against ZZZ? (X
scattergraphs in Excel)? (Come back if this is unfamiliar territory.)

If you do that, you might see interesting features which will bear o
your problem of extrapolation.
XXX against ZZZ shows something like a possible linear correlation, bu
also close to logarithmic. On the evidence, I wouldn't like to have t
call it.

XXX v. YYY is more ?improbable? Best Excel fit is a second-orde
polynomial but not all that convincing.

Al

 

[silky_green]

Thanks Alf! Yeah, not enough data, I know. Also, you're correct about
the YYY data. I think there may have been an error. I may have to add
yet another factor:

XXX YYY ZZZ AAA
100 649 46439 100
200 1531 61384 100
500 5485 89121 100
1000 4117 129452 50

AAA is another column that controls amount of input. So, XXX is
number of users, AAA is amount of input and YYY and ZZZ are data
collected during the user/amount inputs. I'm trying to figure out at
what point, how many XXX and AAA, will I run out of YYY and ZZZ. I'll
try and do the scattergraph as you suggested. If you can think of
anything else, I would really appreciate it.

Thank you again!