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
|
+--------------------------------------------------------------+
| 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 0.5. See CHANGES.1
for earlier versions whose version numbers were the same as release
dates.
Mar 6 2003 to Mar 27 2003 -- version 0.5.008
* made it so you could export a path before doing a fit
* fixed a typo preventing the saving of the bkg data
* "skip" is now a choice (along with guess, set, def) for defining a
parameter. a skip is one you want to keep but do not want to
appear in any way in the ifeffit script
* including a path in a fit via the intrp menu does not redisplay
the intrp display
* for adding a path via the intrp menus, can now choose to just add
or add + jump-to.
* GSD page now displays correctly when the math expression right
click menu is used to define the first parameter
* weed out repeatedly defined parameters before reporting on error
bars and correlations
* added little up and down buttons for re-ordering the list of gds
parameters
* allow user to save multi-column data file for all selected paths
* use labels rather than groups names when saving paths to files
* in after-fit results palette, write the paths in a less surprising
order
* slightly more sensible behavior for state of certain save cascades
in file menu
* in the Paths tree, moved fit, bkg, and resid to branch positions
underneath the data. This will be handy when multiple data set
fits come about.
* popup contextual menus by C-Mouse-3 or Mouse-2 clicking on a list
entry
* allow renaming data, feff, and paths
* implemented get_string from echo area as in athena. use it for
renaming and for getting an R value for excluding/deleting paths
* fixed a bug plotting paths in k-space before a fit
* changes to how plot keys are generated to reflect new ability to
change names
Feb 7 2003 to Feb 28 2003 -- version 0.5.007
* Clicking mouse-3 on a list entry changes the anchor to that entry
and changes the display to the entry but does not change the
selection.
* fixed a bug in deleting a project
* fixed a bug in right clicking on an empty path parameter entry box
* small change to how command line args are handled so that fully
resolved path finds its way into the MRU
Jan 16 2003 to Jan 28 2003 -- version 0.5.006
* implemented the same configuration scheme using tied hashes as in
the recent release of athena
* implemented a similar MRU scheme as in the recent release of
athena
Jan 9 2003 -- version 0.5.005
* very minor (and incomplete) changes to how rcfiles are read when
the prograam starts.
Dec 2 2002 to Dec 13 2002 -- version 0.5.004
* added a "browse" button to the files palette
* added a window_multiplier to the plot parameters from the rc file.
This specifies the height of the window function relative to the
data.
* added a view path file option to the intrp pop-up menu
* removed delay in displaying the dialog for editing math
expressions for many paths
* fixed a problem computing the window function for chi(k) plots for
different k-weights
* added an option to the GDS menu to "reset" all variables, which
means to use the initial guesses for plotting and showing
Nov 14 2002 to Nov 20 2002 -- version 0.5.003
* fixed a bug in parameter verifier which failed to noticed
parameters used that were not defined in the case where no
parameters were defined.
* changed some of the default operational parameter values
* take care to mark guesses for updating when reading in a template
* if previous fit attempt generated error messages, clear those
messages from the files palette before running the current fit.
* deal with chi(k) data files that are not on a uniform grid
* do something sensible if "feff.bin" is selected for a feff calc
rather than "feff.inp"
Oct 25 2002 to Nov 5 2002 -- version 0.5.002
* had to disable truncation of residual arrays to fitting range for
this version to avoid a problem I don't yet understand with
Ifeffit.pm's put::array(). Instead, I am replacing points outside
the fit range with 0 or very small numbers.
* use Config::IniFiles (big changes inside, nothing visible -- I
hope -- to the user)
* allow non-integer degeneracies
* many changes to the bookkeeping issues surrounding paths.
developed a new scheme for assigning and keeping track of path
indeces (i.e. the index for the path and fit commands). this new
scheme is independent both of the N in feffNNNN, of the location
of the path in the list, and of the HList id of the entry. this
has solved numerous problems related to deleting paths and
deleting feff calculations. it also continues and strengthens the
solution to the vanishing-first-path problem mentioned below.
* use the sub from the old Ifeffit::IO for converting a list of
indeces like (3,1,14,5,15,2,13,7,8,6,12) to "1-3,5-8,12-15"
* added MUCH more information to the results palette. this makes it
resemble much more closely a proper log file, la feffit.log
* added some code so that when a project is read in that has the old
pcelem/pcedge buglet, the value for pcedge is interpreted and used
to set pcelem and pcedge.
* added a legend to the feff page explaining the colors in the intrp
box
* resolved an ambiguity as to what value of a parameter should be
used when plotting a path or doing an ff2chi. in short, Artemis
now keeps track of whether a parameter has been altered. if it
has, the new value in the widget is used, if not the value in
ifeffit's memory is used (if it exists) so that guesses and defs
have the most up-to-date values. the flags marking an edited
parameter are unset whenever a plot, fit, or ff2chi happens.
* fixed a bug which triggered a spurious error message for a
pathparam or set/def math expression which started with a minus
sign
* include a check to compare the number of guess parameters to
2*deltak*deltr/pi. Generate an error message and refuse to fit if
bigger.
* corrected some internal problems which prevented duplicated paths,
either by cloning or by re-reading the same feffNNNN.dat file.
* improved the scheme by which the labels displayed in the path list
are assigned to feff paths.
* a potentially fragile change -- changed some details about how
projects get written out and read in to accomodate the changes
related to replicated paths. I have made an effort to correct
prior project files, but there may still be issues.
* new additions to the hints file
* fixed a couple problems determining paths to files, one related to
feffNNNN.dat files, one related to the default project file name
* protect label path param with quotes
* added a query to the Data menu to compute the number of
independent points using Nyquist + 0.
* display top of paths list after a fit or ff2chi
* the right click on word in a path param math expression would not
correctly add a new entry to the GDS list. this is fixed.
* fixed palette so the big text boxes resize correctly as the window
is resized
* new scheme for labeling paths in the plot: F/NNNN where F is the
feff index and NNNN is from feffNNNN.dat. If the path is
duplicated, then the key is F/NNNN:D where D is the duplication
index.
* correctly update titles when switching data files
* use data file name rather than "Data" in the path list and as the
plot key. also use this when constructing suggested filenames for
save_data.
* keyboard plotting shortcuts C-p k, C-p r, and C-p q
* C-a and C-u for selecting/deselecting all paths and data for
plotting
* in intrp box, pathlines change color as mouse passes over. also
added menu functionality to all lines in the box. the header pops
up a menu that replaces the View file buttons, lines that
correspond to paths in the paths list pop up the old menu, lines
that correspond to feffNNNN.dat files that exist but have not been
read in pop up a menu for adding that path, lines that correspond
to feffNNNN.dat files that do not exist (in the -e sense) pop up a
disabled menu that says just that
* removed the View buttons from the feff page, added a view cascade
to the Feff menu
* some cosmetic changes to main page and path page
* added edge energy to intrp header
* save central atom and edge (determined during intrp) to feff
object, use these to set PC parameters for the data object
* do PC FTs using the full phase shift from a selected path. the
menu for selecting the is kept up to date each time the data page
is displayed
* intrp pop-up menu now allows you to select/deselect the path under
the cursor for plotting
* changed text of include checkbutton to contain the path lable from
the paths list
Sep 27 2002 to Oct 24 2002 -- version 0.5.001
* broke source code up into manageable bits
* changed background color to antiquewhite3 and 2 to distinguish
from Athena
* changed text kmin/kmax/Rmin/Rmax to "k-range" and "R-range" to be
consistent with Athena
* deal with capital letters in the feff.inp filename
* enclose all path/file names in quotes in read_data() and path() --
those pesky folders with spaces in their names!
* added a little box for specifying the degeneracy and added the
degen argument to all path() calls. also added items in the Feff
menu for setting all degeneracies to 1 (a la nodegen from feffit)
and resetting them to their feff-computed values
* fixed a bug that prevented the data from being read into ifeffit
is an Athena record was read as the data file in an Artemis
project
* there was a problem where the phase correction variables could get
set such that the edge parameter was something like "cu k" and the
element parameter was "H". I do not know if this was an artemis
problem or an athena problem (I last noticed it in an artemis
project that read an athena k record), but it does not seem to be
there anymore
* fixed the time lag when raising a palette
* fixed the "vanishing first path" problem. it is VERY confusing
trying to keep path indeces straight. rather than doing so, I
decided to just always bring paths up to date for any plot, fit,
or whatever. this is a little less efficient and fills up the
ifeffit buffer with lots of repetitive stuff, but so be it.
* add a button for saving ifeffit buffer to a file. also limit
ifeffit buffer to 1000 lines of text.
Sep 13 2002 -- version 0.5.000
* begin new versioning scheme and begin releasing Athena and Artemis
together
* fixed a minor problem in read_data that was giving a warning
message
* put version number in splashscreen
* if you read in a project file, that file name/path is the default
for saving a project
* can now read in an Athena record -- i.e. data + the relevant
opparams. a check is made to verify that it is a chi(k) record.
* fixed a problem in how title lines were read from data and updated
as the display changes
* added some help text to the plot cards.
* made plotting k^kw the default plotting option in k-space
* use project_header method now in Tools.pm
* Altered Tk::DirSelect to allow title and text
* allow user to save all included paths in one swoop
* intrp pop-up menu now allows you to delete the path under the
cursor
* fixed (I think) the problem with PCFTs by removing the k= argument
to fftf, which triggers a bug in ifeffit in the case of PCFT (but
not otherwise). (Why did I start using "k=" back on 2002.06.20?)
* deleting a project did not get rid of bkg, fit, or res and it did
not erase all those groups from ifeffit memory. now it does.
* fixed a bug that damaged some functionality after deleting a
project (it was caused by re-evaluating %temp in clear_op rather
than resetting individual elements of the hash)
* startup messages run faster (all done after 5.5 sec rather than 9)
* made a temporary fix for an error reading templates which depended
upon the selected path.
* fixed a problem wherein the variable $n_gsd was reset to 0 when a
project was deleted when, in fact, it should always be equal to
the number of entries on the GDS page. This was causing cryptic
failures reading in project files after another was deleted.
* mouse wheel bindings to scrolled text boxes
* most palette text boxes (not journal) are now ROText. also
standard mouse-3 popup was disabled for all text boxes.
* put all three plot options, fit, and ff2chi in the Plot and Data
menus (i.e. all buttons are also menu items)
* "Prefernces" menu is now "Settings" menu
|