how to randomness different loops?
I generate a block of 'strands', each strand has a random factor applied
to its width. How can I create a duplicate of it such that the ends are
fixed, but the points in between are random. The example code the
strands do not change for each iteration of block, which can lead to
repeating pattern when joined up. If I do not use a seed, then the ends
are different for each iteration, and will not align.
if you run one block, with len= 80, then that will show more or less
what I am trying to explain it should look like, apart from the
endpoints of the blocks in the tessellated version.
module strand(ysep,see){ // rand at 20
ty=ysep+10+rands(-2,2,1,see)[0];
by=ysep+rands(-2,2,1,see)[0];
b=[0,ty];
c=[for (j=[1:1:(len/2)-1]) [j*20,ysep+10+rands(-2,2,1)[0] ]];
d= [(len/2)*20,ty];
e=[(len/2)*20,by];
f=[for
(j=[(len/2)+2:1:len])[(len-j+1)*20,ysep++rands(-2,2,1)[0]]];
g=[0,by] ;
p=concat([b],c,[d],[e],f,[g]);
polygon(p);
}
module block(){
for (j=[0:1:ns-1])
strand(sepj,j7);
}
ns=100; // number of strands
sep=18; // strand separation
len= 20; // length of strand
block();
translate([200,0,0])block();
translate([400,0,0])block();
translate([600,0,0])block();
I am not completely following what you are trying to do, but I think
what you are saying is that you want a random wavy line, and then you
want to attach a random wavy line onto the end of it, and another, and
so on, and you want the ends of those lines to connect.
The best answer would seem to be to generate the entire wavy line in one
operation, rather than in segments.
Failing that, it would depend on exactly how you're generating your
waviness. So far, your problem is that your random end point isn't the
same as the random start point on the next one, right?
So, perhaps to point out the obvious, your start point for the next
segment can't be random.
One answer would be to do what model railroaders do when they want to
have individual people build modules for a layout, and want the modules
to interconnect https://www.nmra.org/introduction-layout-modules:
have the interconnection points not be random at all. Assuming that
your waviness is around a zero point, you could have all segments start
and end at the zero point. But if you do an vertical array of these
strands, I suspect that would be a visible pattern.
Another approach would be to plan out the whole line in advance, so that
you know where the start/end points are before you start building
anything else. But mostly that seems equivalent to generating the
entire line in one pass.
So if I'm understanding this correctly, you want the endpoints of strand_1
to be the start points of strand_2?
I'm pretty sure that in order to do this in OpenSCAD you will need to
generate those end points outside of your loop.
One way to do this is to generate an array of starting point rands in
"block()" or "main" and pass "start_point[j]" and "start_point[j+1]" into
strand() also. These would be used by strand as it's start and end points.
You could also generate a list of rands in "main"
SEED=42284;
start_point_array=rands(-2,2,ns+1,SEED);
and just have block() also pass j into strand(),
so then each itteration of strand() uses that j to lookup it's own start
and end points.
Thus Raymond West hast written on Sun, Oct 16, 2022 at 06:20:26PM +0100, and, according to prophecy, it shall come to pass that:
I generate a block of 'strands', each strand has a random factor applied to
its width. How can I create a duplicate of it such that the ends are fixed,
but the points in between are random. The example code the strands do not
change for each iteration of block, which can lead to repeating pattern when
joined up. If I do not use a seed, then the ends are different for each
iteration, and will not align.
if you run one block, with len= 80, then that will show more or less what I
am trying to explain it should look like, apart from the endpoints of the
blocks in the tessellated version.
module strand(ysep,see){ // rand at 20
ty=ysep+10+rands(-2,2,1,see)[0];
by=ysep+rands(-2,2,1,see)[0];
b=[0,ty];
c=[for (j=[1:1:(len/2)-1]) [j*20,ysep+10+rands(-2,2,1)[0] ]];
d= [(len/2)*20,ty];
e=[(len/2)*20,by];
f=[for
(j=[(len/2)+2:1:len])[(len-j+1)*20,ysep++rands(-2,2,1)[0]]];
g=[0,by] ;
p=concat([b],c,[d],[e],f,[g]);
polygon(p);
}
module block(){
for (j=[0:1:ns-1])
strand(sepj,j7);
}
ns=100; // number of strands
sep=18; // strand separation
len= 20; // length of strand
block();
translate([200,0,0])block();
translate([400,0,0])block();
translate([600,0,0])block();
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
Methinks it should be possible to do occasional random droplets of dirty black, hot engine oil splashing against the viewport from behind and running down into a puddle at the bottom. Delicious!
(Sorry, but my formative years were before the anthropocene had been given a name.)
On 10/17/22 22:29, mikeonenine@web.de wrote:
Methinks it should be possible to do occasional random droplets of dirty black, hot engine oil splashing against the viewport from behind and running down into a puddle at the bottom. Delicious!
(Sorry, but my formative years were before the anthropocene had been given a name.)
Mine were about that far back, I made it to 88 back on the 4th.
Cheers, Gene Heskett.
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.
- Louis D. Brandeis
Genes Web page http://geneslinuxbox.net:6309/
gene heskett wrote:
Mine were about that far back, I made it to 88 back on the 4th.
Congrats! I’m only somewhere between 71 and 72 (not quite sure where). Hope I’ll be able to keep up with things like OpenSCAD when I’m 88!
The trick to tile-able noise is to use a seed that incorporates a modulo.
To tile in one direction, something like:
rands(min,max,1, base_seed + mod(position, positions_until_repeat))
I have some sample code which creates a 2D tile-able noise (opposite edges
of heightmap match), in the following gist:
https://gist.github.com/thehans/35aa000d9a833c160a9f93bd3df25edf
(it also does bicubic interpolation between points, but that's another
topic ;) )
Hans
On Mon, Oct 17, 2022 at 11:11 PM mikeonenine@web.de wrote:
gene heskett wrote:
Mine were about that far back, I made it to 88 back on the 4th.
Congrats! I’m only somewhere between 71 and 72 (not quite sure where).
Hope I’ll be able to keep up with things like OpenSCAD when I’m 88!
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
Thanks for the replies, I'll try and show a bit more explanation. I've
extracted the relevant code from my more complex overall solution
attempt, and I've attached it and a svg (the svg produced by the scad
code.)
Looking at the svg, the first block of strands have the end points set
equal, and when a number of blocks are created and joined, the fact that
the endpoints are equal is not apparent The scad script generates that
image. However, I've a suspicion that for certain values of sep, and
range of rands, in the construction of the connected blocks, then the
repeated constant ends would be visible across the group of blocks. I'm
aiming at a general solution, not one that works for certain values.
If you comment out the fixed ty and by, and uncomment the first pair,
(in the strand module) then the blocks have the random end points, but
they do not persist for each block, and gives 'steps' when the blocks
are aligned.
I may be wrong, but it appears that once a seed is generated for random
numbers, it applies to every instant of rands.
If the length of strands is increased to 1000, say, then it shows what I
would like to achieve by joining ten blocks of length 100.
I could generate a list of rands for ty and by, and increment through
when the block is created, but having a choice of random seeds would be
nice.
I may be missing the obvious. While I'm in winge mode, it's an
irritation that importing an svg that has just been exported it arrives
in a different location.
Ray:
I did not read your code, but when you say the "end points set equal",
is the thickness the same, or are both the thickness and the angle the
same? Are you seeing places where the two ends are in the same
location, but the trajectory of the two pieces are so different that the
joint is obvious? Do you need to get the two end points to go at the
same angle?
Jon
On 10/18/2022 8:13 AM, Ray West wrote:
Thanks for the replies, I'll try and show a bit more explanation. I've
extracted the relevant code from my more complex overall solution
attempt, and I've attached it and a svg (the svg produced by the scad
code.)
Looking at the svg, the first block of strands have the end points set
equal, and when a number of blocks are created and joined, the fact
that the endpoints are equal is not apparent The scad script
generates that image. However, I've a suspicion that for certain
values of sep, and range of rands, in the construction of the
connected blocks, then the repeated constant ends would be visible
across the group of blocks. I'm aiming at a general solution, not one
that works for certain values.
If you comment out the fixed ty and by, and uncomment the first pair,
(in the strand module) then the blocks have the random end points, but
they do not persist for each block, and gives 'steps' when the blocks
are aligned.
I may be wrong, but it appears that once a seed is generated for
random numbers, it applies to every instant of rands.
If the length of strands is increased to 1000, say, then it shows what
I would like to achieve by joining ten blocks of length 100.
I could generate a list of rands for ty and by, and increment through
when the block is created, but having a choice of random seeds would
be nice.
I may be missing the obvious. While I'm in winge mode, it's an
irritation that importing an svg that has just been exported it
arrives in a different location.
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
Hi Jon,
The best way to understand what I am trying to explain, is probably to
run the code, and maybe zoom in, or alter the values. Exactly as it is,
I can not see any endpoint artefacts/vertical lines, etc., but I can
envisage that with different values, 'vertical lines/joins' would be
visible.
On 18/10/2022 13:29, jon wrote:
Ray:
I did not read your code, but when you say the "end points set equal",
is the thickness the same, or are both the thickness and the angle the
same? Are you seeing places where the two ends are in the same
location, but the trajectory of the two pieces are so different that
the joint is obvious? Do you need to get the two end points to go at
the same angle?
Jon
On 10/18/2022 8:13 AM, Ray West wrote:
Thanks for the replies, I'll try and show a bit more explanation.
I've extracted the relevant code from my more complex overall
solution attempt, and I've attached it and a svg (the svg produced by
the scad code.)Looking at the svg, the first block of strands have the end points
set equal, and when a number of blocks are created and joined, the
fact that the endpoints are equal is not apparent The scad script
generates that image. However, I've a suspicion that for certain
values of sep, and range of rands, in the construction of the
connected blocks, then the repeated constant ends would be visible
across the group of blocks. I'm aiming at a general solution, not one
that works for certain values.If you comment out the fixed ty and by, and uncomment the first pair,
(in the strand module) then the blocks have the random end points,
but they do not persist for each block, and gives 'steps' when the
blocks are aligned.I may be wrong, but it appears that once a seed is generated for
random numbers, it applies to every instant of rands.If the length of strands is increased to 1000, say, then it shows
what I would like to achieve by joining ten blocks of length 100.I could generate a list of rands for ty and by, and increment through
when the block is created, but having a choice of random seeds would
be nice.I may be missing the obvious. While I'm in winge mode, it's an
irritation that importing an svg that has just been exported it
arrives in a different location.
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org