Skip to main content

evaluation - How to REALLY Hold arguments and capture FullForm?


I'm trying to write an expression deconstructor or FullForm-capturer; might even call it a parser, maybe, but that might be too glorious a word.


I got some great ideas from "Head and everything but...", but have plunged headlong back into the fog of my own limited understanding again, which got even foggier when I tried to understand "...function arguments on a stack...". I'd be most grateful for ideas, advice, and more clarity.


Consider this example: a + b + c. We know its FullForm is Plus[a, b, c], and I can "capture" that FullForm in a List with something like


Cases[{a + b + c}, head_[args___] -> {head, args}]


which produces


{{Plus, a, b, c}}

Now I go get myself all drunk on recursion and think I can write


SetAttributes[capture, HoldAllComplete];
capture[expr_?AtomQ] := {Head@expr, expr};
capture[head_[args___]] := {head, capture /@ {args}};

so that something like capture[a + b + c] gives me fantastic detail


{Plus, {{Symbol, a}, {Symbol, b}, {Symbol, c}}}


or, in stunningly beautiful TraditionalForm


$\left\{\text{Plus},\left( \begin{array}{cc} \text{Symbol} & a \\ \text{Symbol} & b \\ \text{Symbol} & c \\ \end{array} \right)\right\}$


All is well till I try


capture[1 + 2 + 3] ~~> {Integer, 6}

OH NO, it's not Holding inside! What I want, of course, is


$\left\{\text{Plus},\left( \begin{array}{cc} \text{Integer} & 1 \\ \text{Integer} & 2 \\ \text{Integer} & 3 \\ \end{array} \right)\right\}$


My problem is revealed by traceView2, which I got from "... clearest way to represent ..."


Trace of offending call



We see that 1 + 2 + 3 arrives held to AtomQ, as expected, but greedy old evaluator grabs it right there and smooshes it into 6. Everywhere else in the trace, the same thing happens. I don't know an easy way to stop the evaluator inside my capture faction.


It has occurred to me to copy and adapt the code of traceView2 to my application, since it exploits Trace to get ahead of the evaluator, and I will proceed that way unless I receive a cool answer here. In my dreams I imagine something like


SetAttributes[capture, DudeReallyHoldAllCompleteEverywhere];

or


SetAttributes[capture, HoldAllComplete, Levels -> Infinity];

Answer



Ok, what you have here is some classic example of what is called "Evaluation leaks". So, first, the corrected code:


ClearAll[capture]
SetAttributes[capture, HoldAllComplete];

capture[expr_ /; AtomQ[Unevaluated[expr]]] :=
{Head[Unevaluated[expr]], expr};
capture[head_[args___]] := {head, capture /@ Unevaluated[{args}]};

Now, you can inspect it and compare to your version, to see that you had three such "leaks": in the pattern-testing AtomQ (which does not hold its arguments), in the r.h.s. Head[expr], since Head also does not hold them, and, perhaps most importantly, in capture/@{args}, since Map does not hold its arguments either.


So, evaluation leaks are cases when some code which wasn't supposed to evaluate, actually does. Some most frequent situations which lead to evaluation leaks are either when the code is being passed to / anylized by a function which does not hold it properly, or when some parts of held code are extracted during destructuring (e.g. by Cases), not being wrapped in a holding wrapper head.


The standard thing to do to prevent the evaluation once is to wrap your code in Unevaluated, which I did above. To prevent evaluation in a persistent manner, we usually wrap an expression in Hold, HoldComplete, or perhaps some custom holding wrappers with Hold-attributes, but in such cases you have to take care to strip the holding wrapper when no longer needed.


Some more involved examples of parse-like functions which avoid evaluation leaks can be found in this answer (function depends) and also here (function parse). The problem of evaluation leaks is a very frequent one once you start working with pieces of unevaluated code, which is particularly common when writing parser-like functions, macros, and other metaprogramming-related functions. You can look at some references in my recent answer on metaprogramming, for some more links to code where prevention of evaluation leaks was essential.


Comments

Popular posts from this blog

plotting - Plot 4D data with color as 4th dimension

I have a list of 4D data (x position, y position, amplitude, wavelength). I want to plot x, y, and amplitude on a 3D plot and have the color of the points correspond to the wavelength. I have seen many examples using functions to define color but my wavelength cannot be expressed by an analytic function. Is there a simple way to do this? Answer Here a another possible way to visualize 4D data: data = Flatten[Table[{x, y, x^2 + y^2, Sin[x - y]}, {x, -Pi, Pi,Pi/10}, {y,-Pi,Pi, Pi/10}], 1]; You can use the function Point along with VertexColors . Now the points are places using the first three elements and the color is determined by the fourth. In this case I used Hue, but you can use whatever you prefer. Graphics3D[ Point[data[[All, 1 ;; 3]], VertexColors -> Hue /@ data[[All, 4]]], Axes -> True, BoxRatios -> {1, 1, 1/GoldenRatio}]

plotting - Filling between two spheres in SphericalPlot3D

Manipulate[ SphericalPlot3D[{1, 2 - n}, {θ, 0, Pi}, {ϕ, 0, 1.5 Pi}, Mesh -> None, PlotPoints -> 15, PlotRange -> {-2.2, 2.2}], {n, 0, 1}] I cant' seem to be able to make a filling between two spheres. I've already tried the obvious Filling -> {1 -> {2}} but Mathematica doesn't seem to like that option. Is there any easy way around this or ... Answer There is no built-in filling in SphericalPlot3D . One option is to use ParametricPlot3D to draw the surfaces between the two shells: Manipulate[ Show[SphericalPlot3D[{1, 2 - n}, {θ, 0, Pi}, {ϕ, 0, 1.5 Pi}, PlotPoints -> 15, PlotRange -> {-2.2, 2.2}], ParametricPlot3D[{ r {Sin[t] Cos[1.5 Pi], Sin[t] Sin[1.5 Pi], Cos[t]}, r {Sin[t] Cos[0 Pi], Sin[t] Sin[0 Pi], Cos[t]}}, {r, 1, 2 - n}, {t, 0, Pi}, PlotStyle -> Yellow, Mesh -> {2, 15}]], {n, 0, 1}]

plotting - Mathematica: 3D plot based on combined 2D graphs

I have several sigmoidal fits to 3 different datasets, with mean fit predictions plus the 95% confidence limits (not symmetrical around the mean) and the actual data. I would now like to show these different 2D plots projected in 3D as in but then using proper perspective. In the link here they give some solutions to combine the plots using isometric perspective, but I would like to use proper 3 point perspective. Any thoughts? Also any way to show the mean points per time point for each series plus or minus the standard error on the mean would be cool too, either using points+vertical bars, or using spheres plus tubes. Below are some test data and the fit function I am using. Note that I am working on a logit(proportion) scale and that the final vertical scale is Log10(percentage). (* some test data *) data = Table[Null, {i, 4}]; data[[1]] = {{1, -5.8}, {2, -5.4}, {3, -0.8}, {4, -0.2}, {5, 4.6}, {1, -6.4}, {2, -5.6}, {3, -0.7}, {4, 0.04}, {5, 1.0}, {1, -6.8}, {2, -4.7}, {3, -1....