Home My Page Projects Code Snippets Project Openings diderot
Summary Activity Tracker Tasks SCM

SCM Repository

[diderot] Diff of /trunk/TODO
ViewVC logotype

Diff of /trunk/TODO

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1156, Sun May 8 21:20:52 2011 UTC revision 1254, Mon May 23 19:40:55 2011 UTC
# Line 1  Line 1 
1  NOTE: GLK's approximate ranking of 5 most important tagged with  NOTE: GLK's approximate ranking of 8 most important tagged with
2  [GLK:1], [GLK:2], ...  [GLK:1], [GLK:2], ...
3    
4  ========================  ========================
5  SHORT TERM ============= (*needed* for streamlines & tractography)  SHORT TERM ============= (*needed* for streamlines & tractography)
6  ========================  ========================
7    
8  [GLK:1] Add sequence types (needed for evals & evecs)  [GLK:3] Add sequence types (needed for evals & evecs)
9      syntax      syntax
10          types: ty '{' INT '}'          types: ty '{' INT '}'
11          value construction: '{' e1 ',' … ',' en '}'          value construction: '{' e1 ',' … ',' en '}'
12          indexing: e '{' e '}'          indexing: e '{' e '}'
13  [GLK:1] evals & evecs for symmetric tensor[3,3] (requires sequences)  
14    [GLK:4] evals & evecs for symmetric tensor[2,2] and
15    tensor[3,3] (requires sequences)
16    
17  ability to emit/track/record variables into dynamically re-sized  ability to emit/track/record variables into dynamically re-sized
18  runtime buffer  runtime buffer
# Line 18  Line 20 
20  tensor fields: convolution on general tensor images  tensor fields: convolution on general tensor images
21    
22  ========================  ========================
23  SHORT-ISH TERM ========= (to make using Diderot less annoying/slow)  SHORT-ISH TERM ========= (to make using Diderot less annoying to
24  ========================  ========================  program in, and slow to execute)
25    
26    value-numbering optimization [DONE]
27    
28    Allow ".ddro" file extensions in addition to ".diderot"
29    
30    Be able to output values of type tensor[2,2] and tensor[3,3];
31    (currently only scalars & vectors).  Want to add some regression tests
32    based on this and currently can't
33    
34    [GLK:1] Add a clamp function, which takes three arguments; either
35    three scalars:
36      clamp(lo, hi, x)  = max(lo, min(hi, x))
37    or three vectors of the same size:
38      clamp(lo, hi, [x,y])  = [max(lo[0], min(hi[0], x)),
39                               max(lo[1], min(hi[1], y))]
40    This would be useful in many current Diderot programs.
41    One question: clamp(x, lo, hi) is the argument order used in OpenCL
42    and other places, but clamp(lo, hi, x) is much more consistent with
43    lerp(lo, hi, x), hence GLK's preference
44    
45  value-numbering optimization  [GLK:2] Proper handling of stabilize method
46    
47  proper handling of stabilize method  allow "*" to represent "modulate": per-component multiplication of
48    vectors, and vectors only (not tensors of order 2 or higher).  Once
49    sequences are implemented this should be removed: the operation is not
50    invariant WRT basis so it is not a legit vector computation.
51    
52  [GLK:2] Save Diderot output to nrrd, instead of "mip.txt"  implicit type promotion of integers to reals where reals are
53    required (e.g. not exponentiation "^")
54    
55    [GLK:5] Save Diderot output to nrrd, instead of "mip.txt"
56    For grid of strands, save to similarly-shaped array    For grid of strands, save to similarly-shaped array
57    For list of strands, save to long 1-D (or 2-D for non-scalar output) list    For list of strands, save to long 1-D (or 2-D for non-scalar output) list
58    For ragged things (like tractography output), will need to save both    For ragged things (like tractography output), will need to save both
59      complete list of values, as well as list of start indices and lengths      complete list of values, as well as list of start indices and lengths
60      to index into complete list      to index into complete list
61    
62  [GLK:3] Use of Teem's "hest" command-line parser for getting  [GLK:6] Use of Teem's "hest" command-line parser for getting
63  any input variables that are not defined in the source file  any "input" variables that are not defined in the source file.
64    
65  [GLK:4] ability to declare a field so that probe positions are  [GLK:7] ability to declare a field so that probe positions are
66  *always* "inside"; with various ways of mapping the known image values  *always* "inside"; with various ways of mapping the known image values
67  to non-existant index locations.  One possible syntax emphasizes that  to non-existant index locations.  One possible syntax emphasizes that
68  there is a index mapping function that logically precedes convolution:  there is a index mapping function that logically precedes convolution:
69    F = bspln3 ⊛ (img  clamp)    F = bspln3 ⊛ (img ◦ clamp)
70    F = bspln3 ⊛ (img ◦ repeat)    F = bspln3 ⊛ (img ◦ repeat)
71    F = bspln3 ⊛ (img ◦ mirror)    F = bspln3 ⊛ (img ◦ mirror)
72  where "◦" or "∘" is used to indicate function composition  where "◦" or "∘" is used to indicate function composition
73    
 Use ∇⊗ etc. syntax  
     syntax [DONE]  
     typechecking  
     IL and codegen  
   
 Add a clamp function, which takes three arguments; either three scalars:  
   clamp(x, minval, maxval)  = max(minval, min(maxval, x))  
 or three vectors of the same size:  
   clamp([x,y], minvec, maxvec)  = [max(minvec[0], min(maxvec[0], x)),  
                                    max(minvec[1], min(maxvec[1], y))]  
 This would be useful in many current Diderot programs.  
 One question: clamp(x, minval, maxval) is the argument order  
 used in OpenCL and other places, but clamp(minval, maxval, x)  
 would be more consistent with lerp(minout, maxout, x).  
   
74  Level of differentiability in field type should be statement about how  Level of differentiability in field type should be statement about how
75  much differentiation the program *needs*, rather than what the kernel  much differentiation the program *needs*, rather than what the kernel
76  *provides*.  The needed differentiability can be less than or equal to  *provides*.  The needed differentiability can be less than or equal to
77  the provided differentiability.  the provided differentiability.
78    
79    Use ∇⊗ etc. syntax
80        syntax [DONE]
81        typechecking
82        IL and codegen
83    
84  Add type aliases for color types  Add type aliases for color types
85      rgb = real{3}      rgb = real{3}
86      rgba = real{4}      rgba = real{4}
# Line 76  Line 93 
93    
94  "initially" supports lists  "initially" supports lists
95    
96  "initially" supports lists of positions output from  "initially" supports lists of positions output from different
97  different initalization Diderot program  initalization Diderot program (or output from the same program;
98    e.g. using output of iso2d.diderot for one isovalue to seed the input
99    to another invocation of the same program)
100    
101  Communication between strands: they have to be able to learn each  Communication between strands: they have to be able to learn each
102  other's state (at the previous iteration).  Early version of this can  other's state (at the previous iteration).  Early version of this can
# Line 94  Line 113 
113    
114  support for Python interop and GUI  support for Python interop and GUI
115    
116    Allow integer exponentiation ("^2") to apply to square matrices,
117    to represent repeated matrix multiplication
118    
119  Alow X *= Y, X /= Y, X += Y, X -= Y to mean what they do in C,  Alow X *= Y, X /= Y, X += Y, X -= Y to mean what they do in C,
120  provided that X*Y, X/Y, X+Y, X-Y are already supported.  provided that X*Y, X/Y, X+Y, X-Y are already supported.
121  Nearly every Diderot program would be simplified by this.  Nearly every Diderot program would be simplified by this.
# Line 142  Line 164 
164  (but we should only duplicate over the live-range of the result of the  (but we should only duplicate over the live-range of the result of the
165  conditional.  conditional.
166    
167  [GLK:5] Want: non-trivial field expressions & functions:  [GLK:8] Want: non-trivial field expressions & functions.
168    scalar fields from scalar fields F and G:
169      field#0(2)[] X = (sin(F) + 1.0)/2;
170      field#0(2)[] X = F*G;
171    scalar field of vector field magnitude:
172    image(2)[2] Vimg = load(...);    image(2)[2] Vimg = load(...);
173    field#0(2)[] Vlen = |Vimg ⊛ bspln3|;    field#0(2)[] Vlen = |Vimg ⊛ bspln3|;
174  to get a scalar field of vector length, or  field of normalized vectors (for LIC and vector field feature extraction)
175      field#2(2)[2] F = ...
176      field#0(2)[2] V = normalize(F);
177    scalar field of gradient magnitude (for edge detection))
178    field#2(2)[] F = Fimg ⊛ bspln3;    field#2(2)[] F = Fimg ⊛ bspln3;
179    field#0(2)[] Gmag = |∇F|;    field#0(2)[] Gmag = |∇F|;
180  to get a scalar field of gradient magnitude, or  scalar field of squared gradient magnitude (simpler to differentiate):
181    field#2(2)[] F = Fimg ⊛ bspln3;    field#2(2)[] F = Fimg ⊛ bspln3;
182    field#0(2)[] Gmsq = ∇F•∇F;    field#0(2)[] Gmsq = ∇F•∇F;
183  to get a scalar field of squared gradient magnitude, which is simpler  There is value in having these, even if the differentiation of them is
184  to differentiate.  However, there is value in having these, even if  not supported (hence the indication of "field#0" for these above)
 the differentiation of them is not supported (hence the indication  
 of "field#0" for these above)  
   
 Want: ability to apply "normalize" to a field itself, e.g.  
   field#0(2)[2] V = normalize(Vimg ⊛ ctmr);  
 so that V(x) = normalize((Vimg ⊛ ctmr)(x)).  
 Having this would simplify expression of standard LIC method, and  
 would also help express other vector field expressions that arise  
 in vector field feature exraction.  
185    
186  Permit fields composition, especially for warping images by a  Introduce region types (syntax region(d), where d is the dimension of the
187  smooth field of deformation vectors  region.  One useful operator would be
188            dom : field#k(d)[s] -> region(d)
189    Then the inside test could be written as
190            pos ∈ dom(F)
191    We could further extend this approach to allow geometric definitions of
192    regions.  It might also be useful to do inside tests in world space,
193    instead of image space.
194    
195    co- vs contra- index distinction
196    
197    Permit field composition:
198    field#2(3)[3] warp = bspln3 ⊛ warpData;    field#2(3)[3] warp = bspln3 ⊛ warpData;
199    field#2(3)[] F = bspln3 ⊛ img;    field#2(3)[] F = bspln3 ⊛ img;
200    field#2(3)[] Fwarp = F ◦ warp;    field#2(3)[] Fwarp = F ◦ warp;
201  So Fwarp(x) = F(warp(X)).  Chain rule can be used for differentation  So Fwarp(x) = F(warp(X)).  Chain rule can be used for differentation.
202    This will be instrumental for expressing non-rigid registration
203    methods (but those will require co-vs-contra index distinction)
204    
205  Allow the convolution to be specified either as a single 1D kernel  Allow the convolution to be specified either as a single 1D kernel
206  (as we have it now):  (as we have it now):
207    field#2(3)[] F = bspln3 ⊛ img;    field#2(3)[] F = bspln3 ⊛ img;
208  or, as a tensor product of kernels, one for each axis, e.g.  or, as a tensor product of kernels, one for each axis, e.g.
209    field#0(3)[] F = (bspln3 ⊗ bspln3 ⊗ tent) ⊛ img;    field#0(3)[] F = (bspln3 ⊗ bspln3 ⊗ tent) ⊛ img;
210  This is especially important for things like time-varying data, or  This is especially important for things like time-varying fields
211  other multi-dimensional fields where one axis of the domain is very  and the use of scale-space in field visualization: one axis of the
212  different from the rest.  What is very unclear is how, in such cases,  must be convolved with a different kernel during probing.
213  we should notate the gradient, when we only want to differentiate with  What is very unclear is how, in such cases, we should notate the
214  respect to some of the axes.  gradient, when we only want to differentiate with respect to some
215    subset of the axes.  One ambitious idea would be:
216      field#0(3)[] Ft = (bspln3 ⊗ bspln3 ⊗ tent) ⊛ img; // 2D time-varying field
217      field#0(2)[] F = lambda([x,y], Ft([x,y,42.0]))     // restriction to time=42.0
218      vec2 grad = ∇F([x,y]);                             // 2D gradient
219    
220    Tensors of order 3 (e.g. gradients of diffusion tensor fields, or
221    hessians of vector fields) and order 4 (e.g. Hessians of diffusion
222    tensor fields).
223    
224  co- vs contra- index distinction  representation of tensor symmetry
   
 some indication of tensor symmetry  
225  (have to identify the group of index permutations that are symmetries)  (have to identify the group of index permutations that are symmetries)
226    
227  dot works on all tensors  dot works on all tensors
228    
229  outer works on all tensors  outer works on all tensors
230    
231    Help for debugging Diderot programs: need to be able to uniquely
232    identify strands, and for particular strands that are known to behave
233    badly, do something like printf or other logging of their computations
234    and updates.
235    
236    Permit writing dimensionally general code: Have some statement of the
237    dimension of the world "W" (or have it be learned from one particular
238    field of interest), and then able to write "vec" instead of
239    "vec2/vec3", and perhaps "tensor[W,W]" instead of
240    "tensor[2,2]/tensor[3,3]"
241    
242    Traits: all things things that have boilerplate code (especially
243    volume rendering) should be expressed in terms of the unique
244    computational core.  Different kinds of streamline/tractography
245    computation will be another example, as well as particle systems.
246    
247  Einstein summation notation  Einstein summation notation
248    
249  "tensor comprehension" (like list comprehension)  "tensor comprehension" (like list comprehension)
250    
251    Fields coming from different sources of data:
252    * triangular or tetrahedral meshes over 2D or 3D domains (of the
253      source produced by finite-element codes; these will come with their
254      own specialized kinds of reconstruction kernels, called "basis
255      functions" in this context)
256    * Large point clouds, with some radial basis function around each point,
257      which will be tuned by parameters of the point (at least one parameter
258      giving some notion of radius)
259    
260  ======================  ======================
261  BUGS =================  BUGS =================
262  ======================  ======================

Legend:
Removed from v.1156  
changed lines
  Added in v.1254

root@smlnj-gforge.cs.uchicago.edu
ViewVC Help
Powered by ViewVC 1.0.0