Home My Page Projects Code Snippets Project Openings diderot

# SCM Repository

[diderot] Diff of /branches/ein16/TODO
 [diderot] / branches / ein16 / TODO

# Diff of /branches/ein16/TODO

revision 1155, Sun May 8 14:43:30 2011 UTC revision 1257, Tue May 24 18:47:46 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  other SHORT TERM =============  (including needed for LIC)  SHORT TERM ============= (*needed* for streamlines & tractography)
6  ==============================  ========================

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).
7
8  Level of differentiability in field type should be statement about how  Remove CL from compiler
much differentiation the program *needs*, rather than what the kernel
*provides*.  The needed differentiability can be less than or equal to
the provided differentiability.
9
10  [GLK:1] Add sequence types (needed for evals & evecs)  [GLK:3] Add sequence types (needed for evals & evecs)
11      syntax      syntax
12          types: ty '{' INT '}'          types: ty '{' INT '}'
13          value construction: '{' e1 ',' … ',' en '}'          value construction: '{' e1 ',' … ',' en '}'
14          indexing: e '{' e '}'          indexing: e '{' e '}'
15
16  IL support for higher-order tensor values (matrices, etc).  [GLK:4] evals & evecs for symmetric tensor[2,2] and
17      tensor construction [DONE]  tensor[3,3] (requires sequences)
tensor indexing [DONE]
tensor slicing
verify that hessians work correctly [DONE]
18
19  Use ∇⊗ etc. syntax  ability to emit/track/record variables into dynamically re-sized
20      syntax [DONE]  runtime buffer
typechecking
IL and codegen
21
22  test/uninit.diderot:  tensor fields: convolution on general tensor images
documents need for better compiler error messages when output variables
are not initialized; the current messages are very cryptic
23
24  determinant ("det") for tensor[3,3]  ========================
25    SHORT-ISH TERM ========= (to make using Diderot less annoying to
26    ========================  program in, and slow to execute)
27
28  expand trace in mid to low translation  value-numbering optimization [DONE]
29
30  value-numbering optimization  Allow ".ddro" file extensions in addition to ".diderot"
31
32  Add type aliases for color types  Be able to output values of type tensor[2,2] and tensor[3,3];
33      rgb = real{3}  (currently only scalars & vectors).  Want to add some regression tests
34      rgba = real{4}  based on this and currently can't
35
36  ==============================  [GLK:1] Add a clamp function, which takes three arguments; either
37  MEDIUM TERM ================== (including needed for streamlines & tractography)  three scalars:
38  ==============================    clamp(lo, hi, x)  = max(lo, min(hi, x))
39    or three vectors of the same size:
40      clamp(lo, hi, [x,y])  = [max(lo[0], min(hi[0], x)),
41                               max(lo[1], min(hi[1], y))]
42    This would be useful in many current Diderot programs.
43    One question: clamp(x, lo, hi) is the argument order used in OpenCL
44    and other places, but clamp(lo, hi, x) is much more consistent with
45    lerp(lo, hi, x), hence GLK's preference
46
47  [GLK:1] evals & evecs for symmetric tensor[3,3] (requires sequences)  [GLK:2] Proper handling of stabilize method
48
49  [GLK:2] Save Diderot output to nrrd, instead of "mip.txt"  allow "*" to represent "modulate": per-component multiplication of
50    vectors, and vectors only (not tensors of order 2 or higher).  Once
51    sequences are implemented this should be removed: the operation is not
52    invariant WRT basis so it is not a legit vector computation.
53
54    implicit type promotion of integers to reals where reals are
55    required (e.g. not exponentiation "^")
56
57    [GLK:5] Save Diderot output to nrrd, instead of "mip.txt"
58    For grid of strands, save to similarly-shaped array    For grid of strands, save to similarly-shaped array
59    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
60    For ragged things (like tractography output), will need to save both    For ragged things (like tractography output), will need to save both
61      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
62      to index into complete list      to index into complete list
63
64  [GLK:3] Use of Teem's "hest" command-line parser for getting  [GLK:6] Use of Teem's "hest" command-line parser for getting
65  any input variables that are not defined in the source file  any "input" variables that are not defined in the source file.
66
67  [GLK:4] ability to declare a field so that probe positions are  [GLK:7] ability to declare a field so that probe positions are
68  *always* "inside"; with various ways of mapping the known image values  *always* "inside"; with various ways of mapping the known image values
69  to non-existant index locations.  One possible syntax emphasizes that  to non-existant index locations.  One possible syntax emphasizes that
70  there is a index mapping function that logically precedes convolution:  there is a index mapping function that logically precedes convolution:
# Line 76  Line 73
73    F = bspln3 ⊛ (img ◦ mirror)    F = bspln3 ⊛ (img ◦ mirror)
74  where "◦" or "∘" is used to indicate function composition  where "◦" or "∘" is used to indicate function composition
75
76  extend norm (|exp|) to all tensor types [DONE for vectors and matrices]  Level of differentiability in field type should be statement about how
77    much differentiation the program *needs*, rather than what the kernel
78  ability to emit/track/record variables into dynamically re-sized  *provides*.  The needed differentiability can be less than or equal to
79  runtime buffer  the provided differentiability.

Want: allow X *= Y, X /= Y, X += Y, X -= Y to mean what they do in C,
provided that X*Y, X/Y, X+Y, X-Y are already supported.
Nearly every Diderot program would be simplified by this.
80
81  [GLK:5] Want: non-trivial field expressions & functions:  Use ∇⊗ etc. syntax
82    image(2)[2] Vimg = load(...);      syntax [DONE]
83    field#0(2)[] Vlen = |Vimg ⊛ bspln3|;      typechecking
84  to get a scalar field of vector length, or      IL and codegen
field#2(2)[] F = Fimg ⊛ bspln3;
field#0(2)[] Gmag = |∇F|;
to get a scalar field of gradient magnitude, or
field#2(2)[] F = Fimg ⊛ bspln3;
field#0(2)[] Gmsq = ∇F•∇F;
to get a scalar field of squared gradient magnitude, which is simpler
to differentiate.  However, there is value in having these, even if
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.
85
86  tensor fields: convolution on general tensor images  Add type aliases for color types
87        rgb = real{3}
88        rgba = real{4}
89
90  ==============================  ==============================
91  other MEDIUM TERM ============ (needed for particles)  MEDIUM TERM ================== (*needed* for particles)
92  ==============================  ==============================
93
Put small 1-D and 2-D fields, when reconstructed specifically by tent
and when differentiation is not needed, into faster texture buffers.
test/illust-vr.diderot is good example of program that uses multiple
such 1-D fields basically as lookup-table-based function evaluation

94  run-time birth of strands  run-time birth of strands
95
96  "initially" supports lists  "initially" supports lists
97
98  "initially" supports lists of positions output from  "initially" supports lists of positions output from different
99  different initalization Diderot program  initalization Diderot program (or output from the same program;
100    e.g. using output of iso2d.diderot for one isovalue to seed the input
101    to another invocation of the same program)
102
103    Communication between strands: they have to be able to learn each
104    other's state (at the previous iteration).  Early version of this can
105    have the network of neighbors be completely static (for running one
106    strand/pixel image computations).  Later version with strands moving
107    through the domain will require some spatial data structure to
108    optimize discovery of neighbors.
109
110    ============================
111    MEDIUM-ISH TERM ============ (to make Diderot more useful/effective)
112    ============================
113
114  spatial data structure that permits strands' queries of neighbors  Python/ctypes interface to run-time
115
116  proper handling of stabilize method  support for Python interop and GUI
117
118  test/vr-kcomp2.diderot: Add support for code like  Allow integer exponentiation ("^2") to apply to square matrices,
119    to represent repeated matrix multiplication
120
121          (F1 if x else F2)@pos  Alow X *= Y, X /= Y, X += Y, X -= Y to mean what they do in C,
122    provided that X*Y, X/Y, X+Y, X-Y are already supported.
123    Nearly every Diderot program would be simplified by this.
124
125  This will require duplication of the continuation of the conditional  Put small 1-D and 2-D fields, when reconstructed specifically by tent
126  (but we should only duplicate over the live-range of the result of the  and when differentiation is not needed, into faster texture buffers.
127  conditional.  test/illust-vr.diderot is good example of program that uses multiple
128    such 1-D fields basically as lookup-table-based function evaluation
129
130    expand trace in mid to low translation
131
132    extend norm (|exp|) to all tensor types [DONE for vectors and matrices]
133
134    determinant ("det") for tensor[3,3]
135
136  add ":" for tensor dot product (contracts out two indices  add ":" for tensor dot product (contracts out two indices
137  instead of one like •), valid for all pairs of tensors with  instead of one like •), valid for all pairs of tensors with
138  at least two indices  at least two indices
139
140  ==============================  test/uninit.diderot:
141  other MEDIUM TERM ============  documents need for better compiler error messages when output variables
142  ==============================  are not initialized; the current messages are very cryptic
143
144  want: warnings when "D" (reserved for differentiation) is declared as  want: warnings when "D" (reserved for differentiation) is declared as
145  a variable name (get confusing error messages now)  a variable name (get confusing error messages now)
146
support for Python interop and GUI

Python/ctypes interface to run-time

Allow the convolution to be specified either as a single 1D kernel
(as we have it now):
field#2(3)[] F = bspln3 ⊛ img;
or, as a tensor product of kernels, one for each axis, e.g.
field#0(3)[] F = (bspln3 ⊗ bspln3 ⊗ tent) ⊛ img;
This is especially important for things like time-varying data, or
other multi-dimensional fields where one axis of the domain is very
different from the rest.  What is very unclear is how, in such cases,
we should notate the gradient, when we only want to differentiate with
respect to some of the axes.

==============================
LONG TERM ====================
147  ==============================  ==============================
148    LONG TERM ==================== (make Diderot more interesting/attractive from
149    ==============================  a research standpoint)
150
151    IL support for higher-order tensor values (matrices, etc).
152        tensor construction [DONE]
153        tensor indexing [DONE]
154        tensor slicing
155        verify that hessians work correctly [DONE]
156
157  Better handling of variables that determines the scope of a variable  Better handling of variables that determines the scope of a variable
158  based on its actual use, instead of where the user defined it.  So,  based on its actual use, instead of where the user defined it.  So,
# Line 172  Line 160
160  scope.  Also prune out useless variables, which should include field  scope.  Also prune out useless variables, which should include field
161  variables after the translation to mid-il.  variables after the translation to mid-il.
162
163    test/vr-kcomp2.diderot: Add support for code like
164            (F1 if x else F2)@pos
165    This will require duplication of the continuation of the conditional
166    (but we should only duplicate over the live-range of the result of the
167    conditional.
168
169    [GLK:8] Want: non-trivial field expressions & functions.
170    scalar fields from scalar fields F and G:
171      field#0(2)[] X = (sin(F) + 1.0)/2;
172      field#0(2)[] X = F*G;
173    scalar field of vector field magnitude:
175      field#0(2)[] Vlen = |Vimg ⊛ bspln3|;
176    field of normalized vectors (for LIC and vector field feature extraction)
177      field#2(2)[2] F = ...
178      field#0(2)[2] V = normalize(F);
179    scalar field of gradient magnitude (for edge detection))
180      field#2(2)[] F = Fimg ⊛ bspln3;
181      field#0(2)[] Gmag = |∇F|;
182    scalar field of squared gradient magnitude (simpler to differentiate):
183      field#2(2)[] F = Fimg ⊛ bspln3;
184      field#0(2)[] Gmsq = ∇F•∇F;
185    There is value in having these, even if the differentiation of them is
186    not supported (hence the indication of "field#0" for these above)
187
188    Introduce region types (syntax region(d), where d is the dimension of the
189    region.  One useful operator would be
190            dom : field#k(d)[s] -> region(d)
191    Then the inside test could be written as
192            pos ∈ dom(F)
193    We could further extend this approach to allow geometric definitions of
194    regions.  It might also be useful to do inside tests in world space,
196
197  co- vs contra- index distinction  co- vs contra- index distinction
198
199  some indication of tensor symmetry  Permit field composition:
200      field#2(3)[3] warp = bspln3 ⊛ warpData;
201      field#2(3)[] F = bspln3 ⊛ img;
202      field#2(3)[] Fwarp = F ◦ warp;
203    So Fwarp(x) = F(warp(X)).  Chain rule can be used for differentation.
204    This will be instrumental for expressing non-rigid registration
205    methods (but those will require co-vs-contra index distinction)
206
207    Allow the convolution to be specified either as a single 1D kernel
208    (as we have it now):
209      field#2(3)[] F = bspln3 ⊛ img;
210    or, as a tensor product of kernels, one for each axis, e.g.
211      field#0(3)[] F = (bspln3 ⊗ bspln3 ⊗ tent) ⊛ img;
212    This is especially important for things like time-varying fields
213    and the use of scale-space in field visualization: one axis of the
214    must be convolved with a different kernel during probing.
215    What is very unclear is how, in such cases, we should notate the
216    gradient, when we only want to differentiate with respect to some
217    subset of the axes.  One ambitious idea would be:
218      field#0(3)[] Ft = (bspln3 ⊗ bspln3 ⊗ tent) ⊛ img; // 2D time-varying field
219      field#0(2)[] F = lambda([x,y], Ft([x,y,42.0]))     // restriction to time=42.0
221
222    Tensors of order 3 (e.g. gradients of diffusion tensor fields, or
223    hessians of vector fields) and order 4 (e.g. Hessians of diffusion
224    tensor fields).
225
226    representation of tensor symmetry
227  (have to identify the group of index permutations that are symmetries)  (have to identify the group of index permutations that are symmetries)
228
229  dot works on all tensors  dot works on all tensors
230
231  outer works on all tensors  outer works on all tensors
232
233    Help for debugging Diderot programs: need to be able to uniquely
234    identify strands, and for particular strands that are known to behave
235    badly, do something like printf or other logging of their computations
237
238    Permit writing dimensionally general code: Have some statement of the
239    dimension of the world "W" (or have it be learned from one particular
240    field of interest), and then able to write "vec" instead of
241    "vec2/vec3", and perhaps "tensor[W,W]" instead of
242    "tensor[2,2]/tensor[3,3]"
243
244    Traits: all things things that have boilerplate code (especially
245    volume rendering) should be expressed in terms of the unique
246    computational core.  Different kinds of streamline/tractography
247    computation will be another example, as well as particle systems.
248
249  Einstein summation notation  Einstein summation notation
250
251  "tensor comprehension" (like list comprehension)  "tensor comprehension" (like list comprehension)
252
253    Fields coming from different sources of data:
254    * triangular or tetrahedral meshes over 2D or 3D domains (of the
255      source produced by finite-element codes; these will come with their
256      own specialized kinds of reconstruction kernels, called "basis
257      functions" in this context)
258    * Large point clouds, with some radial basis function around each point,
259      which will be tuned by parameters of the point (at least one parameter
260      giving some notion of radius)
261
262  ======================  ======================
263  BUGS =================  BUGS =================
264  ======================  ======================

Legend:
 Removed from v.1155 changed lines Added in v.1257