Skip to content

expected arguments for rgl.triangles

2 messages · Alexander Shenkin, Duncan Murdoch

Original
Alexander Shenkin
# 
Hello all,

I have a set of points in 3D space that represent vertices of a 
non-convex polyhedron.  I would like to plot this polyhedron, and have 
been trying to do so with rgl.triangles, but to no avail.  I imagine I 
don't understand what rgl.triangles expects for arguments.  I have 
constructed the triangles myself, and have the three vertices of each 
triangle.  I am then passing these to rgl.triangles in the x,y & z 
arguments.

I end up with unexpected results (see http://i.imgur.com/BaKfzn7.png). 
Might someone advise me about the structure of arguments that 
rgl.triangles expects (or perhaps a better way to go about this)? 
?rgl.triangles doesn't help much with this particular issue...

Thanks,
Allie


The code, where rows 2,3,4 are one triangle, 5,6,7 the next, etc 
(apologies - don't know how to serialize a dataframe to make it easy to 
suck into one's environment):

rgl.triangles(tri_df[,1],tri_df[,2],tri_df[,3],col=heat.colors(nrow(tri_df)),alpha=.2)

 > tri_df
             x          y   z
2  -2.9970624 -0.1327280 8.0
3  -1.3358202  3.8762849 4.9
4   3.5380065  0.6652143 9.1
5  -1.3358202  3.8762849 4.9
6   1.4644874  3.6145922 5.3
7   3.5380065  0.6652143 9.1
8   1.4644874  3.6145922 5.3
9   1.9606671 -3.8269811 4.5
10  3.5380065  0.6652143 9.1
11  1.9606671 -3.8269811 4.5
12 -3.5337687  3.6772923 3.9
13  3.5380065  0.6652143 9.1
14 -3.5337687  3.6772923 3.9
15 -5.6586500  1.2726665 3.5
16  3.5380065  0.6652143 9.1
17 -5.6586500  1.2726665 3.5
18  3.8327817 -2.8895994 3.4
19  3.5380065  0.6652143 9.1
20  3.8327817 -2.8895994 3.4
21  6.0806691 -0.4852461 2.2
22  3.5380065  0.6652143 9.1
23  6.0806691 -0.4852461 2.2
24 -0.5109254 -6.7807784 0.5
25  3.5380065  0.6652143 9.1
26 -0.5109254 -6.7807784 0.5
27  2.3503422 -6.2742244 1.8
28  3.5380065  0.6652143 9.1
29  2.3503422 -6.2742244 1.8
30  5.4741865 -1.1803740 0.3
31  3.5380065  0.6652143 9.1
32  5.4741865 -1.1803740 0.3
33  5.7888812 -0.3589635 2.7
34  3.5380065  0.6652143 9.1
35  5.7888812 -0.3589635 2.7
36 -3.1939122 -2.4080958 5.4
37  3.5380065  0.6652143 9.1
38 -3.1939122 -2.4080958 5.4
39 -1.4732303 -3.9331403 4.9
40  3.5380065  0.6652143 9.1
41 -1.4732303 -3.9331403 4.9
42 -3.4863074  0.3092904 6.8
43  3.5380065  0.6652143 9.1
44 -3.4863074  0.3092904 6.8
45  0.8019969 -2.2620347 7.2
46  3.5380065  0.6652143 9.1
47  0.8019969 -2.2620347 7.2
48 -2.9970624 -0.1327280 8.0
49  3.5380065  0.6652143 9.1
50 -1.3358202  3.8762849 4.9
51 -2.9970624 -0.1327280 8.0
52  6.4508128 -2.4488802 0.3
53  1.4644874  3.6145922 5.3
54 -1.3358202  3.8762849 4.9
55  6.4508128 -2.4488802 0.3
56  1.9606671 -3.8269811 4.5
57  1.4644874  3.6145922 5.3
58  6.4508128 -2.4488802 0.3
59 -3.5337687  3.6772923 3.9
60  1.9606671 -3.8269811 4.5
61  6.4508128 -2.4488802 0.3
62 -5.6586500  1.2726665 3.5
63 -3.5337687  3.6772923 3.9
64  6.4508128 -2.4488802 0.3
65  3.8327817 -2.8895994 3.4
66 -5.6586500  1.2726665 3.5
67  6.4508128 -2.4488802 0.3
68  6.0806691 -0.4852461 2.2
69  3.8327817 -2.8895994 3.4
70  6.4508128 -2.4488802 0.3
71 -0.5109254 -6.7807784 0.5
72  6.0806691 -0.4852461 2.2
73  6.4508128 -2.4488802 0.3
74  2.3503422 -6.2742244 1.8
75 -0.5109254 -6.7807784 0.5
76  6.4508128 -2.4488802 0.3
77  5.4741865 -1.1803740 0.3
78  2.3503422 -6.2742244 1.8
79  6.4508128 -2.4488802 0.3
80  5.7888812 -0.3589635 2.7
81  5.4741865 -1.1803740 0.3
82  6.4508128 -2.4488802 0.3
83 -3.1939122 -2.4080958 5.4
84  5.7888812 -0.3589635 2.7
85  6.4508128 -2.4488802 0.3
86 -1.4732303 -3.9331403 4.9
87 -3.1939122 -2.4080958 5.4
88  6.4508128 -2.4488802 0.3
89 -3.4863074  0.3092904 6.8
90 -1.4732303 -3.9331403 4.9
91  6.4508128 -2.4488802 0.3
92  0.8019969 -2.2620347 7.2
93 -3.4863074  0.3092904 6.8
94  6.4508128 -2.4488802 0.3
95 -2.9970624 -0.1327280 8.0
96  0.8019969 -2.2620347 7.2
97  6.4508128 -2.4488802 0.3
Duncan Murdoch
#
On 14/11/2014 7:18 AM, Alexander Shenkin wrote:
I can't tell what you think is wrong.   Your arguments look right. 
Here's some general advice:

1.  Don't use rgl.triangles().  Use triangles3d().  This won't affect 
the geometry, but it handles colours and other attributes more rationally.

2.  Think about using a triangle mesh instead of triangles3d().  You 
only need to specify each vertex once, instead of
many times as your current code does.

3.  Put together a simpler example, e.g. two triangles, and see if that 
works.  Then build up from there.

Duncan Murdoch