Sunday, 19 February 2012

odd behaviour of reindex

From: Jens Kaiser
Date: 31 January 2012 02:30


Hi all,
 we encountered an odd behaviour of REINDEX.

Snip form logfile:

 Data line--- reindex HKL (h+l)/2, -k, (h-l)/2
 Data line--- end

 Reflections will be reindexed, and unit cell recalculated

 Reindexing transformation:
      (h' k' l') =  ( h  k  l ) (  1.00000  0.00000  1.00000 )
                                (  0.00000 -1.00000  0.00000 )
                                (  0.50000  0.00000 -0.50000 )

Obviously, the first line of the matrix is not what we intended to
create.

inputting the transformation as HKL h/2+l/2, -k, h/2-l/2
produces the desired result:

 Data line--- reindex HKL h/2+l/2, -k, h/2-l/2
 Data line--- end

 Reflections will be reindexed, and unit cell recalculated

 Reindexing transformation:
      (h' k' l') =  ( h  k  l ) (  0.50000  0.00000  0.50000 )
                                (  0.00000 -1.00000  0.00000 )
                                (  0.50000  0.00000 -0.50000 )



Admittedly, the documentation does not use any brackets in the examples,
but i would expect REINDEX either to throw an error or treat (h+l)/2
like (h-l)/2 but not treat them in the way encountered.

Cheers,

Jens



----------
From: Phil Evans


I had hoped that Pointless could replace Reindex, but doing some test it seems that it also has problems with the syntax of reindex operators. I need to look into this (but perhaps not immediately, sorry)

Phil

----------
From: Eleanor Dodson


Sorry, but  it is a pain writing interpreters - at least it was in Fortran! and once you have one which recognises slashes as divisors, brackets can seem a step too far!

Eleanor

----------
From: Ian Tickle


Jens

Reindex is not treating (h+l)/2 any differently from (h-l)/2, they are
being treated identically, i.e. it is simply ignoring the brackets in
both cases (which is why expanding the expressions without brackets
works)..

So (h+l)/2 is being treated as though you had specified h+l/2 and
(h-l)/2 becomes h-l/2.

Don't forget that the rule for matrix multiplication is row-by-column,
not row-by-row!

Cheers

-- Ian

----------
From: Ian Tickle


Eleanor

Maybe the interpreter should at least throw an error if it encounters
a character (such as a bracket) that it can't interpret, as Jens
implied.

Cheers

-- Ian

----------
From: Phil Evans


Pointless does at least throw an error

----------
From: Eleanor Dodson


Agreed..

E

----------
From: Jens Kaiser


Ian,
 Ah! the old the row is the column depiction of matrices in CCP4 - ha
had forgotten about that! Now at least the output makes sense and the
caveat is to never use brackets.

Thanks to all replies!!

Jens


No comments:

Post a Comment