Appendix: in Dimad, but not yet in the Manual.

Date: Sat, 16 Sep 2000 07:39:04 -0700
From: Roger Servranckx <servranc@home.com>
To: kaltchev@triumf.ca
Cc: Craddock@triumf.ca
Subject: fitting using movement

Dear Dobrin
The introduction of fitting using movement analysis is now done on dimad
in the directory 
~roger/dirnewdim.
It seems to work ok though more tests need to be done.
The example file is in the same directory : testlsq
the corresponding output file is : lsqout

The numeric codes are 701 to 712 for the following variables :
  nux chromx betax alphax etax etapx nuy chromy betay alphay etay etapy

The result is not quite as accurate as with the regular fitting because
we use movement analysis.
We may need to check on the choice of ddelta.
I also use three energies to be able to fit more variables than just the
chromaticities.

Changes in the program are as follows :
Common CTUNE in dimad 09,12(twice),18,24,28
Common FOUT in dimad 09 12(twice) 17 24
Common FITL in dimad 06,07(twice) 09,11,12(three times),13,15,17,27


Subroutine SETOUT in dimad 24
Subroutine Fitmov in dimad 12 
Subroutine fitlsq in dimad 12 (An error was corrected : 780 becomes 760
)
Function FFFCT in dimad09

I think that is all. All changes have been marked DK & RVS Sept 2000 (I
hope!!)

Bye for now
Roger