Subject: RE: MarchingCubeELD question
Date: Wed, 12 Nov 2003 12:25:38 -0500
From: "Michael Hartman" <mhz@engin.umich.edu>
To: "'Michal Husak'" <Michal.Husak@vscht.cz>

Hello Michal,

Thank you for your help.  I have attached a copy of the GSAS density
output .grd.  This is the output of FORGRID by Brian Toby of NIST.  The
data is arranged in a column format with the entries corresponding to

x(0), y(0), z(0)
x(0), y(0), z(1)
...
x(0), y(0), z(number of z steps)
x(0), y(1), z(0)
...
x(number of x steps), y(number of y steps), z(number of z steps)

Therefore the data represent vertical columns that are swept out in y
and then x to cover the map region.

I have another couple of questions.  I am interested in Ettringite which
belongs to a trigonal space group.  From what I can see, the entries
under the CELL definition do not set up the trigonal space group, but
rather this is done by applying a transformation matrix with the TRAN
input.  Is this a proper understanding of how the code handles non-cubic
structures?

Also, the transformation matrix and translation vector seem to apply to
the atomic arrangement relative to the density map information that is
input under the BLOCK sections.  Correct?

What I really need to do is input a trigonal structure and overlay the
corresponding density map.  My current understanding is that in order to
do this I will have to make a density map that has had a rotation matrix
separately applied to bring it into alignment with the underlying
crystalline structure and put in zeros outside of the unit cell.  Is
this the easiest way to approach this problem?

Best regards,
Mike

Michael R. Hartman
Nuclear Engineering and Radiological Sciences Department
U n i v e r s i t y    o f     M i c h i g a n
2301 Bonisteel Boulevard
Ann Arbor, MI 48109-2100
Phone: (734) 764-0055
Fax:     (734) 936-1571


-----Original Message-----
From: Michal Husak [mailto:Michal.Husak@vscht.cz]
Sent: Tuesday, November 11, 2003 10:49 AM
To: Michael Hartman
Subject: Re: MarchingCubeELD question

Hi Michael

>I am very interested in using your program to visualize my GSAS data in

>3D.  I have my fourier map files output into an ascii array and Im
>attempting to put them into an input form suitable for MarchingCubeELD.
I
>am basing my input on the InputDescription.txt file that you supplied
>along with the software.
>
>I have a couple of question to ensure that I am interpreting the input
>correctly.
>
>What is the transformation matrix and translation vector in the TRAN
part
>of the input and how do I calculate the proper values?

This is a matrix with is used for atom position transformation into the
coordinate system of the calculated ELD map.

If you multiply the fractional coordinates of atoms by this matrix + add

the translation part of the matrix
to the data you will have atom position in the Cartesian coordinate
system
in with the ELD map is calulated ...
The reversion of this matrix is internally used if you calculate from a
peak position an atom position in
fractional data ...
This matrix should be generated by the code with had generated your ASCI

file ...
For cartesina ELD map output from the source program, this matrix could
be
set to identity:
1 0 0
0 1 0
0 0 1
0 0 0
This is done in WinGX software during output generation for MCE ...

>If I wanted to add deuterium to the program, is this done simply by
>placing appropriate text into the Atompar2.txt file?

It should work ...

Can you, please, send me some sample GSAS data in 3D with some input
format
description ?
I will add direct input from GSAS to my code, because it probably have
sense ...


Michal

    ---------------------------------------------------------------------
               Name: TEST.grd
   TEST.grd    Type: application/x-unknown-content-type-Photoshop.Gradients
           Encoding: 7bit
