###################################################################
LATEST EMAIL UPDATE: 02 Sep 2013
###################################################################

I just updated the csurf tarfiles here:

   http://www.cogsci.ucsd.edu/~sereno/.tmp/dist/csurf0.8-mac-130902.tgz
   http://www.cogsci.ucsd.edu/~sereno/.tmp/dist/csurf0.8-linux-130902.tgz

This update includes and initial version of
using retinotopic region-growing to find
borders.  This was the original reason I
started hacking on csurf again in Sept 2010...

For each vertex in a label, you pick a seed,
sort other vertices in by distance from the
seed, then start growing nearest-first, while
compiling a list of RF positions.  If any
new vertex-to-be-added is close to anybody
in the current list, mark that vertex as a possible
border.

If you want to play around with it, the
instructions are in:

    Help -> Calc Visual Fieldsign, Borders

Use the GROWBORD button next to the
FIELDSIGN button.  I put the controls
on the fieldsign panel since both methods
need an eccen and a polar file.  If you tick the
Show ... checkbutton, you can see every 20th
fill (N.B.: requires a .sphere).

I made it write out two .w files, both
with a _k infix.

The first one, dumped into the input eccen
directory chosen for calculuations (*not* the
fieldsign directory), is just a wfile of the
raw counts of the number of times each vertex
experienced a regression to an already-seen
RF (usu. >2000 for real borders).

The second one, dumped into the companion
polar directory used for calculations, is the
same but multiplied by the following fraction

         num seeds
    --------------------
    num vertices tested

This bring the count number down to around
1.0 for a little more convenient displaying.
The BRy color scale is best because you
can adjust where the bright yellow appears.

It might work even for single dimension data
like tonotopy.  The next project is to write
something similar in 3D for finding maps
in the pulvinar.

Detailed notes of what I've changed in
the last two updates are below.

cheers,
marty

##############
UPDATE DETAILS
##############

### UPDATE: 02 Sep 2013
--lib/help/csurf/{setupfield,setuprender}
--csurf: fieldsign panel sorts complex wfiles to top, Show Progress tick
--borders.tcl: optional live display
--borders.tcl: swap_val_val2 and write val2 to polardir
--tksurfer.c: floodfill_retin_patches check openglwindow, cnt*seeds/tested->val2
--mk0: include borders.tcl in dist,inst
--lib/tcl/help/tkmedit/{bwslope,bwmid}
--lib/tcl/borders.tcl: new fs-make-like script to run floodfill_retin_patches
--csurf: revert0.9->0.8 in csurf.dat header to match tarfiles (TODO:newvers:1.0)
--csurf: borders output dumped into eccen and polar dirs used, *not* fsdir
--csurf: add GROWBORD/parms to run floodfill_retin_patches to fieldsign panel
--tksurfer.c: can run grow borders without openglwindow w/o error
--lib/help/tksurfer/swapstatval (explain overload below, document complex stack)
--tksurfer.tcl: overload swap stat/val with swap complex stack for middle-click
--csurf: expand the width of the fieldsign listbox dropdowns
--tksurfer.tcl: also look for scriptsdir *.lut, add dropdown divider, ignore it
--tksurfer.c: write_annoted_vertices, 0.0 for val: sav fill_annot_holes'd label
--lib/help/tksurfer/{label,label_clear,label_read,smooth,smooth_steps,rgb}
--tksurfer.tcl: overload smoothsteps: "valholes" (vs. num) -> fill_val_holes
--tksurfer.tcl: overload smoothsteps: "labelholes" (vs. num) -> fill_annot_holes
--tksurfer.c: clear_vertex_marks=>clear_vertex_marks_list, restore_ripflag enum
--tksurfer.c: floodfill_retin_patches: retinotopic region growing
--tksurfer.c: fill_{annot,val}_holes: fill single vertex voids in annot or val
--tksurfer.tcl: if read MGH lab + complexflag, ask reset disp (cpx,col,fmid,etc)
--tksurfer.c: toggle off statmaskampflag around disk/rect scalebars to see them
--tksurfer.tcl: read_and_color_labeled_vertices => read_label_to_annot_using_col
--tksurfer.c: read_and_color_labeled_vertices => read_label_to_annot_using_col

### UPDATE: 22 Aug 2013
...




