Date:         Wed, 16 Jul 2003 14:04:50 -0400
Reply-To:     Mike Fay <rmf4@CDC.GOV>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Mike Fay <rmf4@CDC.GOV>
Subject:      Use SAS to fit this curve?

Hi folks,

I'm an amateur user of SAS, and even worse with statistics (laugh).
Recently we charted some descriptive data that I'd like to fit to a curve,
but am unsure how to ask SAS to do this. Or more especially, is there a way
to ask SAS to search through the different fit possibilities and report on
which are the best fit?

The data itself is appended. It shows the number of U.S. hazardous waste
sites at which N or more chemicals are found in an exposure pathway. For
example, the "n_chems = 2, n_sites = 588" observation means that two (or
more) chemicals are found in an exposure pathway at 588 sites. (The
universe N is 1706 -- many sites don't have a single exposure pathway.)
(This data is draft and subject to revision as the generating algorithm is
refined; do not cite or quote.)

If I play around with several possible fits, such as exponential or
logarithmic, the best R2 is from an exponential fit. But it appears to be
unable to cope with the steep drop-off and then levelling off, even though
the curve seems pretty darn smooth to the eye. In other words, although I
get an R2 of .955 in Excel's very simple "put a fit on this data"
functionality, it obviously misses especially the beginning of the curve.
And I don't know what else to do if Excel's missing it... but I'm willing
to bet SAS (v9) can do a lot more than Excel  ;)

Anyone is also welcome to point out any other math/stats software that
might exist, particularly for the Windows environment. And the simpler
(VERY simple) to operate, the better, lol.

TIA if you can help!

   Mike, Atlanta

***

input N_Chems N_Sites;

cards;

 1 743
 2 588
 3 475
 4 390
 5 321
 6 266
 7 214
 8 185
 9 161
 10 136
 11 111
 12 102
 13 87
 14 74
 15 69
 16 61
 17 56
 18 52
 19 47
 20 46
 21 44
 22 39
 23 36
 24 33
 25 29
 26 26
 28 24
 29 22
 30 20
 31 19
 32 18
 37 16
 38 15
 39 13
 40 11
 41 10
 42 9
 48 7
 49 6
 51 5
 56 4
 58 3
 62 2

;