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

SCM Repository

[diderot] Diff of /trunk/doc/probe/paper.tex
ViewVC logotype

Diff of /trunk/doc/probe/paper.tex

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

revision 347, Tue Sep 21 15:31:49 2010 UTC revision 348, Wed Sep 22 15:52:48 2010 UTC
# Line 311  Line 311 
311  \label{fig:3d-probe-code-opencl}  \label{fig:3d-probe-code-opencl}
312  \end{figure}  \end{figure}
313    
314    \section{Generalizing to derivative fields}
315    To generalize the probe operation to work on derivative fields (\eg{}, $(\nabla^k F)\mkw{@}\vecp$),
316    we represent the $\nabla^k$ operation as a $k$-order tensor of partial derivative operators.
317    To simplify the presentation, we will restrict the discussion to $3$-dimensional space,
318    but it easily generalizes.
319    In 3D, a partial-derivative operator has the form
320    \begin{displaymath}
321      \frac{\partial}{\partial{}x^i \partial{}y^j \partial{}z^k}
322    \end{displaymath}%
323    where $i$, $j$, and $k$ are the number of levels of differentiation in the $x$, $y$,
324    and $z$-axes (respectively).
325    The normal convention is to omit explicit mention of axes with zero levels of differentiation,
326    but we will be explicit here.
327    We use $\mathbf{Id}^n$ to represent the identity operator in $n$-dimensions (i.e., the
328    one with zero=-levels of differentiation in all axes).
329    
330    The product of two partial-derivative operators is formed by summing the exponents;
331    for example in 3D we have
332    \begin{displaymath}
333      \frac{\partial}{\partial{}x^i \partial{}y^j \partial{}z^k}
334      \frac{\partial}{\partial{}x^{i'} \partial{}y^{j'} \partial{}z^{k'}} =
335      \frac{\partial}{\partial{}x^{i+i'} \partial{}y^{j+j'} \partial{}z^{k+k'}}
336    \end{displaymath}%
337    The last bit of notation that we need is the $n$-vector of partial-derivative
338    operators $\delta^n$, in which the $i$th element has one level of differentiation in the $i$th
339    axis (and zero in all other axes).
340    For example,
341    \begin{displaymath}
342      \delta^3 = \left[
343        \frac{\partial}{\partial{}x^1 \partial{}y^0 \partial{}z^0} \quad
344        \frac{\partial}{\partial{}x^0 \partial{}y^1 \partial{}z^0} \quad
345        \frac{\partial}{\partial{}x^0 \partial{}y^0 \partial{}z^1}
346      \right]
347    \end{displaymath}%
348    For $k > 0$ levels of differentiation in $n$-dimensional space, we have
349    \begin{displaymath}
350      \nabla^k = \left\{\begin{array}{ll}
351        \mathbf{Id}^n & \text{when $k = 0$} \\
352        \delta^n \otimes \nabla^{k-1} & \text{when $k > 0$} \\
353      \end{array}\right.
354    \end{displaymath}%
355    Consider, for example, the case where $k = 2$.
356    \begin{displaymath}
357      \nabla^2 = \delta^3 \otimes \delta^3 = \left[\begin{array}{ccc}
358        \frac{\partial}{\partial{}x^2 \partial{}y^0 \partial{}z^0}
359        & \frac{\partial}{\partial{}x^1 \partial{}y^1 \partial{}z^0}
360        & \frac{\partial}{\partial{}x^1 \partial{}y^0 \partial{}z^1} \\
361        \frac{\partial}{\partial{}x^1 \partial{}y^1 \partial{}z^0}
362        & \frac{\partial}{\partial{}x^0 \partial{}y^2 \partial{}z^0}
363        & \frac{\partial}{\partial{}x^0 \partial{}y^1 \partial{}z^1} \\
364        \frac{\partial}{\partial{}x^1 \partial{}y^0 \partial{}z^1}
365        & \frac{\partial}{\partial{}x^0 \partial{}y^1 \partial{}z^1}
366        & \frac{\partial}{\partial{}x^0 \partial{}y^0 \partial{}z^2} \\
367      \end{array}\right]
368    \end{displaymath}%
369    
370    When we have a probe of a derivative field $(\nabla^k F)\mkw{@}\vecp$, we will generate code for
371    each partial derivative operator $\frac{\partial}{\partial{}x^i\partial{}y^j\partial{}z^k}$
372    separately.
373    If $F = V\circledast{}h$, then we will use $h^i(x) h^j(y) h^k(z)$ as the interpolation
374    coefficients, where $h^i$ is the $i$th derivative of $h$.
375    
376  \section{Probing a 3D derivative field}  \section{Probing a 3D derivative field}
377  We next consider the case of probing the derivative of a scalar field $F = V\circledast{}h$, where $s$ is the support  We next consider the case of probing the derivative of a scalar field $F = V\circledast{}h$, where $s$ is the support
378  of $h$.  of $h$.

Legend:
Removed from v.347  
changed lines
  Added in v.348

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