Added atan2 test.

Fixed major illust-vr bug that made images too dark. 

and in 3D

updated to show bug

new bug

bug in 2x2 matrix vector multiply?

test for 2x2 eigensolve

  Added comment about checking output

fix unu recipe

fix comments

added header comment

test program for evecs and evals, based on spectral decomposition

now using stabilize

  Use += etc.

now using stabilize!

  Added test of eigenvalues

  simple tests

  Added a couple of test programs

  simplify example

Updated instructions for processing output.

Removed embedded pointers on both the host and gpu side

  update for new type rules

  merging changes from trunk

  merging changes from trunk

  merging changes from trunk

  misc

  Tracking language change

  Split uses of ∇ into scalar ∇ and tensor ∇⊗.

  Minor tweaks

  Added clamp and assignment operators.

  Working on command-line option support

Rolled zslice2.sh back to the previous version.

Added the code for generating OpenCL

  Change "-log" to "--log"

  Minor fixes

  Comment out programs that have syntax issues or that use unimplemented features

  Remove test/Makefile.in, since it is not useful anymore

  Working on debugging value numbering

  Tracking trunk changes

  Switch to using die

  Port updates from trunk

scoping bug fixed; computation of halfway dot product simplified

  remove BUG comment for a fixed bug

now have pi!

this test program has served its purpose

list syntax for initialization

matrix subtraction works

  Added pie as a float literal

simpler way of getting correct subtraction

fix until 3x3 matrix subtraction works

fix until 3x3 matrix subtraction works

this test program has served its purpose

3x3 matrix subtraction is addition

is this not the way to do an if else?

slightly more logical expression of curvature computation

new convolution usage

new convolution usage

slightly simpler

weird bug

design of txf means inside() no longer neede

new 2D rainbow colormap

new 2D rainbow colormap, simpler field definition

new 2D rainbow colormap

vector field convolution works now

  Remove BUG comment

combining image loading and field convolution seems to work better now

combining image loading and field convolution seems to work better now

problems negating a vector-value probe result

  Removed "BUG" comments for things that are working

  Uncomment tests

numbering of remaining problems

subtle bug in trying to create two fields with one-line expressions

simpler field creation

bug in field expressions

forgot that the new name for this is vr-curv-quad.diderot

this test program has served its purpose

image loading and field creation can happen in one expression now

still exhibits scoping bug

added output process info

still have problems with field expressions

this test program as served its purpose

gradient and normal computation shouldn't be done if there is no opacity

  Bug fix: syntax error when attempting to define 1D array of strands

cant seem to have 1D set of strands

cant seem to have 1D set of strands

fixed dataset use, and removed needless normalize

removed needless normalize

how is this not already committed?

demo of true and false

demo of true and false

now finds three isocontours instead of just one, and uses data inside unit square

  Added note about bug

typos and tweaks

accidental commit of usage of non-committed data

previous commit was accidental

previous commit was accidental

previous commit was accidental

only need to screen against 1/gmsq failing

another demo of isosurfacing

  Fixed bug with using nans

  Fixed subtracting negative literal bug

how checking both values, gradients, and hessians

for debugging index-to-world gradient transforms in 2D and 3D

better documenting of nan bug

demo of 2D isocontour sampling

demo of 2D isocontour sampling

this test program as served its purpose

  Fixed issue with normalize on 2D vectors

variable rename

now goes both directions

now goes both directions

normalize() does not work on vec2

demo of line integral convolution

demo of line integral convolution

for debugging expressions involving fields

documents want of better error msg for unitialized variable

  adding test program from trunk

new probe syntax

tweaks

new probe syntax

tweaks

rename

seemingly finished demo of curvature-based transfer function for volume rendering

demo of curvature computation

this test program has served its purpose

added post-processing info

now a good demo of simple volume rendering

  Change syntax for field probing.

new disturbing bug

  updating examples

specular highlight is now more correct: not modulated by material color

documenting bug about expressing negative hessian

  removed BUG comment that is no longer true

this test program has served its purpose

vr-kcomp is for scoping bug, vr-kcomp2 is for fields in conditional expressions

this test program has served its purpose

this test program has served its purpose

new bug, and escalating no-redeclaration want

  Added comment about processing output

  Switch zslice-rgb to tent to make debugging easier

test program for probing vector images

have new curvature test program

no longer needed

removing unused rayDir

waiting on vector convolution

back to using trace()

for curvature demos

  Tweaked example for testing purposes

bug in C code for mat x mat

now using ^ for pow

  Some debugging and added "^" as power operator.

  Get rid of unused extra-code.c file

bug in dot?

  Remove some HEYs

new test programs

hessians now implemented

  Fixed syntax error.

  Added local field variables for testing purposes

removed extraneous initial declaration of val and norm

possible bug with field variables

small simplification; removed grad variable

didn't mean to commit last version

added need for something akin to convolution on vector images for RGB transfer functions

  Fixed E_Slice exception bug in vr-lite-cam.diderot

correct place to set transp = 0.0

stabilize code block not working

made specular highlight more correct, and other camera,shading tweaks

bspln5 not working

camera tweaks, and more consistent use of normalize

specular highlight more correct now, but new bug

now using normalize and new dot

now using normalize and new dot

now using normalize and new dot

now using normalize and new dot

used lerp where possible, other cosmetic changes

this test program has served its purpose

removed HEY about fixed bug, and now with lerp

  Switching examples to newer language mechanisms

making commented out code evoke the bug

this test program has served its purpose

this testing program has served its purpose

now with lerp

  Use lerp

bug

  merge diffs in zslice2

can't yet use c1tent

added a once-differentiable tent kernel called c1tent

new possible bug

new possible bug

now show bug

new bug

new bug

  Removed workaround in camera code

  Tracking bug fixes

demos 2D convolution, when it is implemented

now using tan()

did I never commit this?

tweaks

non-trivial volume renderer

moved camera stuff that isnt per-ray into global decs

now with indication of whats not boilerplate

initial pass at camera code

  Added -m64 flag to C compiles and fixed some warnings.

  workaround until stabilize methods are properly implemented

new volume rendering test

tweak

possible bug

  Remove adding extra C code

not everyone has . in their path

tweaked background value

tweaked background value

  minor tweaks.

  Adding test script for zslice-grad example

  Removed workaround for always stabilizing

  fix bug in test program

  test on hand data

  map infs to 0 (instead of -1)

tweaked

simple test of gradient 

  Added test scripts

  Added comment

  Use tent filter

  Added unu stage to handle infs in output

correcting affine transform to index space of dataset

  Fixed calculation of rr and cc

another for debugging

new test program, for generating an image of the recon kernel

new more questions

hadn't finished generalizing the 200 image size 

test to show two possible bugs

  Test driver

  Added rules for generating the png file

  Fixes to makefile

  Added image decoding to Makefile

  Working on missing pieces of code generation

  Debugging vr-MIP code

  Working on build scripts for Diderot test programs

  change "actor" to "strand" in test cases

  Creating a branch to explore a pure CFG IL

  Changed typing rules for tensor construction.  The convention is now slow-to-fast
  for tensor dimensions.

  Add more type tests

  Added type test example

  Working on C version of MIP

  more work on MIP example

  Working on hand-translated example.

  Working on runtime

  Working on Diderot runtime

  Working on C version of runtime

  Working on vr-mip in C example

  Upated URL in header to diderot-language.cs.uchicago.edu

  Working on a hand translation of vr-MIP.diderot to C code.

  Use the dummy transfer-function data for the vr-phong test

  Separate tests with "===="

  Added a script to run the compiler over a bunch of examples

  Use conditional-expression syntax

  Remove uneeded parens

  Adding file

  removed stabilize command from stabilize method

some usage tips

added use of stabilize code block

fixing output value creating

int -> real

white space change

  updates to test code

initial pass at volume rendering lite

new vfrhand data

new vfrhand data

Updated verision of the mip code that appears to be working fine.

Updated verision of the mip code that appears to be working fine.

Added the OpenCL outlined in the probe paper

Added the OpenCL outlined in the probe paper

  Changed kernels to have 4 pieces

  Fixed makefile

Added the opencl version of the probe code.

  Using Unicode operators

  Switch to infix notation for convolution

  Added infix convolve syntax

  Made transfer file an input

  Remove Makefile, since it is generated from Makefile.in

  Update makefile

  Factor out image index computation

  fixed typo

  Debugging

new camera positions

added camera info for vox1 rendering

  Add Makefile.in to test/MIP

Test Case now produces output that allows for viewing it as a png.

Updated the mip test case

  Added inside test and default value for stepSz input

Added the Opencl files for the MIP test case.

  Update examples with output variables

tweaked params

better params

removed tabs, added example values for txs.nrrd MIP renderings

  Update example to current syntax

  Removed type comments, since that information has been collected

  Workaround for compiler bug

  Fixed type errors in example

  Added initialization

  Fixed type errors

  Added new syntax for tensor-value construction and more work on typechecker.

  Fixed type error

  Minor edits

added annotations about types of operations

simple diffusion tensor tractography example

fixed bug in ray direction computation

fixed bug in ray direction computation

txf is 1-D field of 4-tuples

  Convert to new syntax

1-D txf as field

added a plausible phong-lighting volume renderer

new example

  Added intialization syntax

  Parser hacking

  Added some more code

  Create directory for test programs