Athena normalization and deglitching bugs
Hi Bruce, I figured I'd post this to the Ifeffit list so that others could be alerted to similar problems, but I'll send you an Athena project file exhibiting this problem separately, so that everyone's in-boxes don't get cluttered with attachments. I'm using Athena 0.8.028 on Windows XP. First of all, on any file, including the copper data that comes with Athena, there is a bug. To see it, go to the deglitch screen, and change Emax by picking a point off the plot. Then replot. From now on, the green "tolerance" lines extend from Emax up to the end of the plot, rather than from Emin to Emax. This seems to affect only the screen display (i.e. it still deglitches the intended range), but it does make it harder to see what will get deglitched. The second behavior is best revealed with the project I am sending you separately. I am not positive it is a bug, but it doesn't match what I would expect. In the project I am giving you, the data is very glitchy (lots of Bragg peaks). Go down to the group labeled SBT/B Bi LIII Merge and plot in E-space with the post-edge line. Does it look right to you? It looks like the normalization line is considerably higher than it should be given the normalization range chosen, with only a few glitches above it. Is that something I don't understand about the way the post-edge line is chosen, or is it displaying the line wrong, or is it calculating it wrong? Thanks--the new features in Artemis are great, by the way. --Scott Calvin Sarah Lawrence College
On Saturday 29 May 2004 02:43 pm, Scott Calvin wrote: SC> First of all, on any file, including the copper data that comes with SC> Athena, there is a bug. To see it, go to the deglitch screen, and SC> change Emax by picking a point off the plot. Then replot. From now SC> on, the green "tolerance" lines extend from Emax up to the end of the SC> plot, rather than from Emin to Emax. This seems to affect only the SC> screen display (i.e. it still deglitches the intended range), but it SC> does make it harder to see what will get deglitched. OK. I'll look into this and put it on the to do list. SC> In the project I am giving you, the data is very glitchy (lots of SC> Bragg peaks). Go down to the group labeled SBT/B Bi LIII Merge and SC> plot in E-space with the post-edge line. Does it look right to you? SC> It looks like the normalization line is considerably higher than it SC> should be given the normalization range chosen, with only a few SC> glitches above it. Is that something I don't understand about the way SC> the post-edge line is chosen, or is it displaying the line wrong, or SC> is it calculating it wrong? I looked at your data and I don't believe anything is wrong. The data after the big white line is somewhat concave downward, thus you would expect the post edge line to be similarly concave downward. However, you chose a value for the lower end of the normalization range that was well up into the white line. Having a fairly large number of points in the normalization range and in the white line biased the regression (and it IS a regression and not a fit) such that the post edge line was, as you say, "above the data". When I chose a lower end of the normalization range that was completely beyond the white line, the post edge line went right through the data as expected. B -- Bruce Ravel ----------------------------------- ravel@phys.washington.edu Code 6134, Building 3, Room 405 Naval Research Laboratory phone: (1) 202 767 2268 Washington DC 20375, USA fax: (1) 202 767 4642 NRL Synchrotron Radiation Consortium (NRL-SRC) Beamlines X11a, X11b, X23b National Synchrotron Light Source Brookhaven National Laboratory, Upton, NY 11973 My homepage: http://feff.phys.washington.edu/~ravel EXAFS software: http://feff.phys.washington.edu/~ravel/software/exafs/
participants (2)
-
Bruce Ravel -
Scott Calvin