File: CHANGES.old.3

package info (click to toggle)
horae 071~svn536-1
  • links: PTS
  • area: contrib
  • in suites: wheezy
  • size: 12,996 kB
  • sloc: perl: 67,215; lisp: 744; sh: 78; makefile: 76; ansic: 35
file content (258 lines) | stat: -rw-r--r-- 13,302 bytes parent folder | download | duplicates (4)
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