学习Asymptote时的一些问题

abababa

hejoseph 发表于 2023-4-4 17:36
我这里没问题，不知道你那里什么问题。ASY嵌入文档的应该是my-1_0.pdf这个文件吧。
如果你只是插图完全可 ...

下面的代码my.tex：

1. \documentclass[12pt]{article}
2. \usepackage{amsmath,bm,yhmath,amssymb,etoolbox,esvect,stmaryrd,cite,enumitem,extarrows,mathtools,ifthen}
3. \usepackage[amsmath, thmmarks]{ntheorem}
4. {
5.         \theoremstyle{nonumberplain}
6.         \theoremheaderfont{\indent\bfseries}
7.         \theorembodyfont{\normalfont}
8.         \theoremsymbol{\ensuremath{\Box}}
9.         \newtheorem{proof}{证明}
10. }
11. \usepackage{pgf,tikz}
12. \usepackage[inline]{asymptote}
13. \usepackage{float}
14. \usepackage{pgfplots,wrapfig}
15. \usepackage{graphicx}
16. \usepackage{subfigure}
17. \usepackage{tkz-euclide}
19. \usetikzlibrary{arrows,automata,positioning,matrix,calc,intersections,shapes.geometric}
20. \begin{document}
21. \begin{asy}
22. settings.outformat = "pdf";
23. settings.prc = false;
24. size(5cm,0);
25. import three;
26. draw(unitsphere);
27. //axes3("$x$","$y$","$z$",Arrow3);
28. \end{asy}
29. try.
30. \end{document}

Copy the Code

我先xelatex my.tex，出现一个警告：

1. Package asymptote Warning: file `my-1.tex' not found on input line 31.

Copy the Code

我觉得是还没编译asy文件的原因。之后用asy *.asy，出现下面的错误：

1. shipout3(prefix,f,preview ? nativeformat() : format,

Copy the Code

这个错误之前一直存在，我以为升级版本就好了，但并没有。按照39楼链接里的说明，把settings.outformat = "pdf";里的pdf改成html，这时重复上面的编译，asy *.asy之后，出现了一个my-1.html，这个网页文件里就有那个图。这时再用xelatex my.tex，期望是把那个图弄进my.pdf里，但是出现警告：

1. Package asymptote Warning: file `my-1.tex' not found on input line 31

Copy the Code

图也没有进到my.pdf里。

而之前的平面图，我没有设置outformat，它就是在当前文件夹里生成了一堆图片的pdf，最后进到总的my.pdf里。

我想既然学了这个能画立体图的asymptote，就试着画几个简单的。

abababa

abababa 发表于 2023-4-4 18:35
下面的代码my.tex：

发现加上这一句settings.render=0;就好了，也不用什么outformat之类的，新的代码如下：

2. \begin{asy}
3. settings.render=0;
4. import three;
5. import graph3;
6. size(5cm,0);
7. draw(unitsphere);
8. \end{asy}

Copy the Code

虽然画出来的那个球是全黑的，可能还是有问题，但我感觉单纯地画线应该没问题了。

abababa

对于3D的，过一点作一条直线的垂线，垂足要怎么画？例如：

1. \begin{asy}
2. settings.render=0;
3. import three;
4. import graph3;
5. size(5cm,0);
7. triple A1 = (1,sqrt(1/3),sqrt(8/3)), A3 = (2,0,0), A4 = (1,sqrt(3),0);
8. \end{asy}

Copy the Code

想画过A1垂直于A3A4的直线的垂足。

我查了three.asy，有几个project的，弄出来都不是。

hejoseph

关于render=0的问题，你可以试试运行下这个asy代码：

正常情况下应该输出下图

render=0会输出下图，这是因为没法正确消隐

abababa

hejoseph 发表于 2023-4-6 14:14
关于render=0的问题，你可以试试运行下这个asy代码：
size(200);
import graph3;

shipout3(prefix,f,preview ? nativeformat() : format,

上面这个错误，好像和显卡有关，我昨天发给网友，他那里也是这个问题，没能弄出pdf来，但生成html的那个都是好的。我搜索这个，发现还是有人遇到相同的问题：
linux.debian.bugs.dist.narkive.com/CzpDhAUs/b … te-error-in-shipout3

abababa

abababa 发表于 2023-4-5 19:25
对于3D的，过一点作一条直线的垂线，垂足要怎么画？例如：

我知道那个怎么算，但是想弄个一般的，现在点是用triple定义的，直线怎么定义不知道。
点$(x_0,y_0,z_0)$到点$(x_1,y_1,z_1), (x_2,y_2,z_2)$所成直线的垂足坐标为
\[(k(x_2-x_1)+x_1, k(y_2-y_1)+y_1, k(z_2-z_1)+z_1)\]

其中
\[k=\frac{(x_1-x_0)(x_2-x_1)+(y_1-y_0)(y_2-y_1)+(z_1-z_0)(z_2-z_1)}{(x_2-x_1)^2+(y_2-y_1)^2+(z_2-z_1)^2}\]

hbghlyj

abababa 发表于 2023-3-27 11:24
这个在我这里运行，出现以下问题：
shipout3(prefix,f,preview ? nativeformat() : format,

若本机总是不行, 可以使用Asymptote Web Application [output选择pdf]
如果PDF需要嵌入PRC可以使用Asymptote server的command-line interface

1. curl --data-binary 'import teapot;' 'asymptote.ualberta.ca:10007?f=pdf&prc' -o teapot.pdf

Copy the Code

ASY-Forum

hbghlyj

abababa 发表于 2023-4-3 14:27
搜索到两个类似的：

根据John Bowman用下面的命令编译

1. ./configure --disable-gc
2. make clean
3. make CFLAGS=-g

Copy the Code

用gdb --args asy 1.asy检查

1. Starting program: /usr/local/bin/asy 1.asy
2. [Thread debugging using libthread_db enabled]
3. Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
5. Program received signal SIGSEGV, Segmentation fault.
6. 0x00005555558f2066 in GC_find_limit_with_bound (p=0x555555ac04a8 "",
7.     up=up@entry=0, bound=bound@entry=0x0) at os_dep.c:1018
8. 1018                    GC_noop1((word)(*result));

Copy the Code

用valgrind检查

1. valgrind --track-origins=yes asy test

Copy the Code

输出有很多. 最后几行是

1. ==35368==
2. ==35368== Conditional jump or move depends on uninitialised value(s)
3. ==35368==    at 0x49D2CA: GC_find_header (headers.c:32)
4. ==35368==    by 0x4A186C: GC_mark_and_push_stack (mark.c:1487)
5. ==35368==    by 0x4A1931: GC_push_all_eager (mark.c:1589)
6. ==35368==    by 0x4AF158: GC_with_callee_saves_pushed (mach_dep.c:335)
7. ==35368==    by 0x4A259C: GC_push_regs_and_stack (mark_rts.c:853)
8. ==35368==    by 0x4A259C: GC_push_roots (mark_rts.c:926)
9. ==35368==    by 0x4A0F67: GC_mark_some (mark.c:388)
10. ==35368==    by 0x498004: GC_stopped_mark (alloc.c:880)
11. ==35368==    by 0x498FBE: GC_try_to_collect_inner (alloc.c:626)
12. ==35368==    by 0x499ECC: GC_collect_or_expand (alloc.c:1678)
13. ==35368==    by 0x49A214: GC_allocobj (alloc.c:1784)
14. ==35368==    by 0x49E09E: GC_generic_malloc_inner (malloc.c:194)
15. ==35368==    by 0x49E2C0: GC_generic_malloc (malloc.c:254)
16. ==35368==  Uninitialised value was created by a stack allocation
17. ==35368==    at 0x4AF0C1: GC_with_callee_saves_pushed (mach_dep.c:229)
18. ==35368==
19. ==35368==
20. ==35368== More than 100 errors detected.  Subsequent errors
21. ==35368== will still be recorded, but in less detail than before.
22. ==35368== Use of uninitialised value of size 8
23. ==35368==    at 0x49D2F7: GC_find_header (headers.c:33)
24. ==35368==    by 0x4A186C: GC_mark_and_push_stack (mark.c:1487)
25. ==35368==    by 0x4A1931: GC_push_all_eager (mark.c:1589)
26. ==35368==    by 0x4AF158: GC_with_callee_saves_pushed (mach_dep.c:335)
27. ==35368==    by 0x4A259C: GC_push_regs_and_stack (mark_rts.c:853)
28. ==35368==    by 0x4A259C: GC_push_roots (mark_rts.c:926)
29. ==35368==    by 0x4A0F67: GC_mark_some (mark.c:388)
30. ==35368==    by 0x498004: GC_stopped_mark (alloc.c:880)
31. ==35368==    by 0x498FBE: GC_try_to_collect_inner (alloc.c:626)
32. ==35368==    by 0x499ECC: GC_collect_or_expand (alloc.c:1678)
33. ==35368==    by 0x49A214: GC_allocobj (alloc.c:1784)
34. ==35368==    by 0x49E09E: GC_generic_malloc_inner (malloc.c:194)
35. ==35368==    by 0x49E2C0: GC_generic_malloc (malloc.c:254)
36. ==35368==  Uninitialised value was created by a stack allocation
37. ==35368==    at 0x4AF0C1: GC_with_callee_saves_pushed (mach_dep.c:229)
38. ==35368==
39. ==35437==
40. ==35437== HEAP SUMMARY:
41. ==35437==     in use at exit: 3,281,028 bytes in 15,961 blocks
42. ==35437==   total heap usage: 303,368 allocs, 287,407 frees, 78,911,078 bytes allocated
43. ==35437==
44. ==35437== LEAK SUMMARY:
45. ==35437==    definitely lost: 2,725,919 bytes in 6,671 blocks
46. ==35437==    indirectly lost: 360 bytes in 3 blocks
47. ==35437==      possibly lost: 0 bytes in 0 blocks
48. ==35437==    still reachable: 554,749 bytes in 9,287 blocks
49. ==35437==         suppressed: 0 bytes in 0 blocks
50. ==35437== Rerun with --leak-check=full to see details of leaked memory
51. ==35437==
52. ==35437== For lists of detected and suppressed errors, rerun with: -s
53. ==35437== ERROR SUMMARY: 640245 errors from 101 contexts (suppressed: 0 from 0)
54. ==35368==
55. ==35368== HEAP SUMMARY:
56. ==35368==     in use at exit: 1,050,411 bytes in 11,466 blocks
57. ==35368==   total heap usage: 210,055 allocs, 198,589 frees, 8,857,023 bytes allocated
58. ==35368==
59. ==35368== LEAK SUMMARY:
60. ==35368==    definitely lost: 853,760 bytes in 6,670 blocks
61. ==35368==    indirectly lost: 0 bytes in 0 blocks
62. ==35368==      possibly lost: 0 bytes in 0 blocks
63. ==35368==    still reachable: 196,651 bytes in 4,796 blocks
64. ==35368==         suppressed: 0 bytes in 0 blocks
65. ==35368== Rerun with --leak-check=full to see details of leaked memory
66. ==35368==
67. ==35368== For lists of detected and suppressed errors, rerun with: -s
68. ==35368== ERROR SUMMARY: 640245 errors from 101 contexts (suppressed: 0 from 0)

Copy the Code

没有懂

abababa

hbghlyj 发表于 2023-4-7 07:53
若本机总是不行, 可以使用Asymptote Web Application [output选择pdf]
如果PDF需要嵌入PRC可以使用Asympto ...

不想用那么多工具，要不是为了画圆锥曲线方便点，我就不学这个asymptote了，用tkz-euclide就很好。

hbghlyj

abababa 发表于 2023-4-6 08:39
上面这个错误，好像和显卡有关

我昨天这个问题：
sourceforge.net/p/asymptote/discussion/409349/thread/ca9f0b6278/
John Bowman told me to upgrade graphic drivers:

OpenGL GLSL version 1.40 is very old (circa 2009). For full functionality (e.g. transparency) you will need to upgrade to GLSL 4.3 or later (which dates back to 2013). You should upgrade your drivers.

hbghlyj

abababa 发表于 2023-4-6 08:39
好像和显卡有关...

编译 Vulkan 时出错
汇编消息：

1. {standard input}: Assembler messages:
2. {standard input}:310129: Warning: end of file not at end of a line; newline inserted
3. {standard input}:310173: Error: unknown pseudo-op: `.lvl261'
4. {standard input}: Error: open CFI at the end of file; missing .cfi_endproc directive
5. cc: fatal error: Killed signal terminated program cc1

Copy the Code

嗯，根据 stackoverflow.com/a/25584112，看起来这可能是构建期间内存不足的情况？需要终止其他进程以释放内存。
重新编译多次，终于成功了。

hbghlyj

abababa 发表于 2023-4-6 08:39
好像和显卡有关...

我在 Centos 8 上的错误消息是类似的。
昨天发到Asymptote Discussion
sourceforge.net/p/asymptote/discussion/409349/thread/ca9f0b6278/

John Bowman - 6 hours ago
Under your platform there is a problem with OSMesa initialization. CentOS tends to use very old software. CentOS 8.5.2111 was released sometime ago, in November 2021 and CentOS was subsequently discontinued. Try upgrading to a newer OS.

需要升级操作系统。

hbghlyj

abababa 发表于 2023-4-6 08:39
shipout3(prefix,f,preview ? nativeformat() : format,

上面这个错误，好像和显卡有关，我昨天发给网友 ...

升级到Ubuntu 22.04.2 LTS后编译freeglut和asymptote，根据此讨论设置 gs 路径，尝试渲染3D图形，报错

    shipout3(prefix,f,preview ? nativeformat() : format,

            ^

/usr/local/share/asymptote/three.asy: 2896.13: runtime: to support onscreen OpenGL rendering; please install the glut library, then ./configure; make

因为我已经根据官方文档编译安装了freeglut 3.2.1，所以只需要安装libglm-dev

1. apt-get install libglm-dev

Copy the Code

上述错误消失了。出现新错误：

freeglut (asy): failed to open display ''

Putty设置“X display location=:0”后，“xeyes”和“xclock”运行成功

出现新错误：

X Error of failed request:  GLXUnsupportedPrivateRequest

  Major opcode of failed request:  143 (GLX)

  Minor opcode of failed request:  17 (X_GLXVendorPrivateWithReply)

  Serial number of failed request:  24

  Current serial number in output stream:  24

hbghlyj

Floating-point unit 浮点运算器
A floating-point unit is a part of a computer system specially designed to carry out operations on floating-point numbers. Typical operations are addition, subtraction, multiplication, division, and square root.

FP exceptions GNU libc
The IEEE 754 standard defines five exceptions that can occur during a calculation. Each corresponds to a particular sort of error, such as overflow.

abababa

hbghlyj 发表于 2023-4-14 02:25
试一下
the -mask command-line option  Mask fpu exceptions

加上这个-mask，编译那个最简单的球的那个，就是draw(unitsphere);，这个出来了，而且也是有点立体感的那个球，不是像平面图那样的。其它的还不知道会不会有问题。

hbghlyj

abababa 发表于 2023-4-15 12:42
这个出来了

可以试一下47# hejoseph的代码看输出是否和左图一致.
可以把-mask保存在config.asy中, 就不必每次都加-mask

abababa

hbghlyj 发表于 2023-4-15 19:46
可以试一下47# hejoseph的代码看输出是否和左图一致.
可以把-mask保存在config.asy中, 就不必每次都加-ma ...

47楼的那个，仅用import graph3;是正常的，前面加一个import three;就出问题了。

hbghlyj

abababa 发表于 2023-4-15 13:55
47楼的那个，仅用import graph3;是正常的，前面加一个import three;就出问题了。

试 asymptote.ualberta.ca 仅用import graph3;或import three;是正常的，一并用也正常

abababa

hbghlyj 发表于 2023-4-15 21:07
试 http://asymptote.ualberta.ca 仅用或是正常的，一并用也正常

我这里后来也没问题了，不知道怎么回事，代码我就改了这一句，但是后来手动把那些.asy,.pre等文件都删除了。

hbghlyj

Asymptote forum
Could you run without -mask in the debugger (e.g. gdb) and tell me which library call in glrender.cc the floating point exception is generating the floating point exception? Once we identify the line number we can locally apply the mask like this (there might be more than one place):

1. fpu_trap(false); // Work around FE_INVALID
2. // offending call goes here
3. fpu_trap(settings::trap());

Copy the Code