1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259
|
+--------------------------------------------------------------+
| Change log for Artemis |
| http://leonardo.phys.washington.edu/~ravel/software/exafs |
+--------------------------------------------------------------+
| Bruce Ravel |
| ravel@phys.washington.edu |
+--------------------------------------------------------------+
Begin change log for version numbers starting at 0.6. See CHANGES.2
for the 0.5 releases. See CHANGES.1 for earlier versions whose
version numbers were the same as release dates.
December 2 2003 -- December ?? 2003 0.6.011 (never released, go to 0.7.000)
* make all toplevels in the same class (and common with athena) so
that they group in the taskbar
* correct a situation where spurious path parameter math expressions
may be left in Ifeffit's memory, leading to incorrect evaluation
of a path (one way this might have happened was to define a
pathparam math expression, evaluate the path, delete the math
expression entirely, then reevaluate the path. in that case, the
previous math expression would have been left in Ifeffit's memory,
leading to an incorrect evaluation of that path)
* added several Echo messages specific to certain actions
* cloning a path now halves the degeneracy of the current and cloned
paths
* using the right click menus on a path parameter math expression to
define a new variable on the GDS page now inserts "0" as the math
expression if it is a newly created variable
* the "Add a new variable" button now focusses on the name of the
new variable
* change the page layouts on the GDS, feff, and path views to use a
frame and the packer. this gives more consistent layout on
different screens/font sizes
* added a Messages tab so that errors and warnings issued by Artemis
have a place to go without overwirting the contents of the Files
tab.
* several improvements to the Edit math expression dialog
* write initial guesses to the log file
* made the window a bit narrower and made size of views a global
parameter so it's easier for me to change
* found several places where the flag for inquiring about saving the
project was not being set when it should have been
* when importing feff paths, the choice for importing the first N
paths is now configurable
* added Scott's ZnO example to the distro. removed the Pu and
NiC4N4 examples
* fixed a logic problem in one of the comments in the fitting script
generation which misidentified a normal fit as a difference fit
November 28 2003 -- December 3 2003 0.6.010
* fixed the color of the scrollbar of the prefs description
* added line styles for data, fit, and parts
* disable log viewer when prefs dialog is shown
October 30 2003 -- November 25 2003 0.6.009
* bound &quit_artemis as the delete protocol for the main window
* prompt for verification when clearing variables or converting all
guesses to sets
* clear the Files tab before every fit (so that warnings from
previous fits aren't still reported there)
* disable plot and include data buttons if discarding data brings
the project down to a single data set, also toggle including of
first data set if it is excluded from fit
* several appearence problems uncovered in Tk 804.025
* fixed a problem contructing command string for an fft using a feff
path for the phase correction
* fixed a problem constructing suggested file names for res and diff
output files
* ask ifeffit about its limits on number of paths and such
September 4 2003 -- October 27 2003 -- version 0.6.008
* fixed a problem reporting fit results when a path parameter
evaluates to NaN, resulting in a string of *'s in ifeffit path
report. artemis now handles this gracefully and issues an error
message useful for diagnosing the problem.
* when saving results to a log file, default file name is based on
project file name, if it exists
* made the intrp box in the APA a bit smaller
* added an echo bar message to the "Reset all variables" option in
the GDS menu
* added an echo area message for mouse-over in intrp box
* Clicking the "include path" button now issues an echo area
message.
* Can now specify a path to be ifeffit's current path so that after
the fit the def parameters are evaluated in terms of that path.
This info is saved to the project file.
* Moved the "Show" submenu from Edit to it's own menu.
* Plot running r-factors in k, R, or q. Put a cascade in the Data
menu.
* Log viewing utility, with reports and plots + averages and
Einstein fits. In Edit menu.
* added an xbm icon so the unix windowmanager toolbars/panels/icon
boxes/docks/whatever have something pretty to display
* post explanation of problem when a fit is aborted due to
parenthesis mismatch
* can now convert all guesses to sets in GDS menu
* disable right click menus for the column heads on the GDS view
* modified many keyboard shortcuts to align more closely with common
UI choices, e.g. C-o for open data, C-p for print, and so on.
zoom is now C-= and unzoom is now C--. The key chains for
plotting in k, R, or q are now C-; k, C-; r, and C-; q
(i.e. control-semicolon)
* grab previous and grab next in GDS right click menu
* projectbar underneath menu bar for displaying path+name of current
project + configuration option for displaying file, project title,
or nothing
* closing project now asks if you want to save before closing.
* right click to define a new param with Stay option, vhanges view
on GDS grid without showing GDS grid. also do an update before
moving when clicking the "Add a parameter" button.
* split path() commands into separate lines for each path
parameter. this helps to avoid very long lines caused by long
file names and/or long math expressions
* include error checking for
-- number of paths used
-- number of variables used
* deal with upper/lower case feffnnnn.dat file names. this is an
issue when transferring a project between windows and unix
* fixed bug where a null paths entry would be created if the pcpath
parameter for a data group points at a nonexistant path
* fixed a bug where a fit would abort if a path for a data set NOT
included in the fit does not exist
* added the ability to delete data sets (but not the first one --
use change data, instead)
June 20 2003 -- August 20 2003 -- version 0.6.007
* save and restore plot features in project file
* handle Mac terminated files transparently
* handle long filenames transparently
* added a right click menu to the numbers on the GDS view with the
following two features + a way to clear that variable
* added a way to move GDS params up and down the list more flexibly
than the little up and down buttons
* added a way to ask Artemis where a parameter is used
* can now choose which data set to plot once fit is finished
* fixed a problem managing guess variables that lead to inconsistent
fit results. the problem was an erase() line that was too long
for ifeffit. the solution was to break it up into many shorter
erase()'s
* difference fitting is implemented. the diff spectrum can be made
from an arbitrary set of paths.
* preferences implemented
* can now restore default values for operational params
* added a feature for automated generation of parameters. it's not
super-brilliant, but it should give a decent starting point,
particularly for first shell fits.
* fixed a bug counting data sets in an MDS fit and some data sets
are excluded from the fit.
* figured out how to explicitly set backpage and inactive colors in
the NoteBook widget
* added a dialog at read data time asking if this is to be a
multiple data set fit or a change of data sets. renew data in the
Data menu circumvents this
* fixed a bug where title lines get lost when a data set was changed
* when changing paths to feff, update the headers, reff, etc for
each path + fix a bug when on feff view when changing paths
June 18 2003 -- June 19 2003 -- version 0.6.006
* Fixed a bug handling the is_rec flag
* perl/Tk 804.024 is pickier that earlier versions about minus signs
in front of method arguments, such as width and height, so put in
the dash wherever needed
May 24 2003 -- June 16 2003 -- version 0.6.005
* Fixed bugs in save_all_paths and save_selected
* better column labels (short_descriptor method) and title lines
for save_selected. also use plotting k-weight for r and q
* fixed major problems in saving column data for individual groups.
also, data in r or q is now saved using the current k-weight for
plotting
* also use plotting k-weight for save_all_paths
* fixed a bug identifying the parent in the situation where the
include path menu is used while the feff calcualtion is current
* added dphase to the path parameters list
* added $config{meta}{extpp} and the "extended path parameters"
(dpahse and (k|amp|phase)_array). by default, the extended pps
are hidden. when the config variable is true, they are unhidden
and added to the end of the list. a checkbutton is in the paths
menu for unhiding them. when hidden, they are not used in the fit
* Now give the user the option of carrying on even if certain errors
are found in the math expressions. This is a modal dialog box:
Continue or Abort.
* fixed problem in Project Assistant where the same feff calc could
not be imported for more than one data set (which is useful for a
multiple-temperature fit)
* removed Control-c keyseq, replaced it with Control-period
* allow restraints to be defined on the GDS page and use them as
such in the fit. (all restraints are applied to the final
data_set) Added an example of using a restraint to the tarball.
* put epsilon_k on the data view. put an item in the data menu for
reporting epsilon_k
* allow saving selected as k-weighted chi(k)
* erase delta_ scalars before the fit
* added some more options to the right click menu for words in the
path parameter math expressions. added skip + the option not to
jump to the GDS view
* verify that an data file for import is actually a data file
* fixed a bug where athena chi(k) records were only read correctly
as the first data set
* try to make sure that the plot options have sensible values
* constrain the fitting space variable to be the same for every data
set (like cormin)
* added a right click menu for most op params
* added a "Save as ..." to the file menu. Made Control-s save
without prompting if a default project filename already exists.
Project saving now writes a .bak backup file.
May 2 2003 -- May 18 2003 -- version 0.6.004
* use features of ifeffit 1.0077
* refuse to run without Ifeffit 1.0077 or higher installed
* new syntax for multiple k weight fits
* fix a minor bug in project saving (only save titles for data
groups)
* make use of the list of ill-posed variables reported by ifeffit
when writing results of a fit
* fixed a bug displaying keys containing commas (by enclosing the
key string in quotes)
* report reff+dr for each path in the results
* fixed a bug where skip variables were not saved to the project
file
* made foreground text of anchored, unincluded paths the brownish
color rather than the black
* do a simple check of math expressions to verify that parens match
* reporsition Data and Paths list when a new data file is imported
April 24 2003 -- May 2 2003 -- version 0.6.003
* add a confirmation dialog to project deletion. also move the
delete project choice farther away from the save project in the
File menu
* check that files are writable before writing
* fixed a bug in path selection where the selection rule was not
sticking to the current feff calc in situations where that makes
sense.
April 22 2003 -- April 23 2003 -- version 0.6.002
* move rc and mru files to the .horae directory
* fixed a bug whereby the first path may be assigned index=0
* fixed a bug involving the C-n (rename_this) keyboard shortcut when
the shortcut is done on a feff calculation
* fixed a bug regarding how titles get saved to project files
* exclude a few data-group parameters from the project file, one of
which was causing trouble in the Safe box evaluation of the
parameter array when the project was read in
April 21 2003 -- version 0.6.001
* Multiple data sets + many changes to accommodate this
* multiple k-weights
* Artemis Project Assistent
* removed templates (see version 2002.08.19)
* report Nidp in echo area whenever kmin,kmax,rmin,rmax are changed
* fixed a bug in up/down buttons and skip variables
* always compute residual in k-space
* fixed a bug running ff2chi which resulted in bad low-r Fourier
components
* changed layout of widgets controlling the plotting options, also
make widgets for plotting k-weight separate from fitting k-weight
|