Dear Cora,

If you want your wRp factor be meaningful you should keep your standard
deviations in the data file. 

After saying that, coming back to the real world, it seems that your problem
is related with the format that CONVERT is using to write down the file in
STD format. In the STD format of GSAS it is assumed that sigma(I)=sqrt(I),
the only way in which this is true is that the intensity is the true number
of counts, thus a positive integer.

If you still want to use CONVERT to do this you should re-scale your
intensities to avoid truncation of non-integer numbers, in your case each
intensity by 100 and you'll get integers. Another way of doing this (without
writing the whole GSAS file again with the re-scaled intensities) could
directly be deleting all the decimal points in the intensity part of the
file (just a global find-replace). I guess CONVERT can deal with lines with
less than 80 characters. If not I suggest you using an EXCELL spreadsheet or
similar program to format your data in the needed way.

If you are planning to ignore the standard deviations I would suggest you to
look for a scale factor that approximately makes sqrt(I)~esd(I) for the most
of your intensities you can, in that way at least the R factors will not be
completely meaningles. Usualy if you rescale your data by X you can easily
get X from esd^2/I.

Good luck.
Leo

 


Dr. Leopoldo Suescun
Postodoctoral Appointee
Materials Science Division - Bldg 223 - Rm D217 
Argonne National Laboratory
9700 S. Cass Ave., Argonne, IL 60439
Phone: 1 (630) 252 9760
Fax: 1 (630) 252 7777
URL: http://www.msd.anl.gov/groups/nxrs/personnel/suescun/index.html
-----Original Message-----
From: Cora Lind [mailto:[EMAIL PROTECTED] 
Sent: Wednesday, June 28, 2006 4:46 PM
To: [email protected]

Hello everybody!

I am trying to convert a GSAS ESD file to a GSAS STD file (to use in
FullProf). I successfully used the program "Convert" to do so, but when I
plotted my data, I realized that Convert erased all my numbers after the
decimal point, leading to very crude patterns with very obvious steps (this
is high-pressure data, so for higher pressures, the overall intensity
differences are not all that high). Does anybody know a simple solution to
this problem, like a different conversion program, or a way to get Convert
to keep the 2 decimal places?

Cora

*******************************************
Dr. Cora Lind
WO 2262

Department of Chemistry, MS 602
University of Toledo
Toledo, OH 43606-3390

Phone:  (419) 530-1505 (W)      (419) 472-8364 (H)      (607) 592-4650 (Cell
- when 
traveling)
Fax:            (419) 530-4033 (W)      (419) 472-8364 (H)

e-mail: [EMAIL PROTECTED]
*******************************************

BEGIN:VCARD
VERSION:2.1
N:Suescun;Leopoldo
FN:Leopoldo Suescun ([EMAIL PROTECTED])
ORG:Argonne National Laboratory - Materials Science Division
TITLE:Postdoctoral Appointee
TEL;WORK;VOICE:+1 (630) 252-9760
TEL;HOME;VOICE:+1 (630) 910-1562
TEL;WORK;FAX:+1 (630) 252-7777
ADR;WORK:;;9700 S. Cass Ave.;Argonne;Illinois;60439;USA
LABEL;WORK;ENCODING=QUOTED-PRINTABLE:9700 S. Cass Ave.=0D=0AArgonne, Illinois 60439=0D=0AUSA
URL;WORK:http://www.msd.anl.gov/groups/nxrs/people/suescun.html
EMAIL;PREF;INTERNET:[EMAIL PROTECTED]
REV:20050630T182807Z
END:VCARD

Reply via email to