calculate a value with a switch from 1 to 0

Programming Questions

clankill3r

calculate a value with a switch from 1 to 0

in Programming Questions • 1 year ago

We have a path,

the length is normalized where the start of the path is 0.0 and the end is 1.0

Now let's say we can close the path, then the start is the same as the end, 1.0 / 0.0.

Every ellipse on the path is a anchor, the ones in red hold a value called weight (shown as w=6, w=.., w=15).

The ones in green also have a weight but this will depend on the 2 neighbors with a red ellipse.

I'm able to call the weight value for anchors on the dotted line (no anchors show for that in the picture).

But i can't get it done between the 0.0 and 1.0.

Showing code is not really a option but pseudo code or just explanation is welcome.

I guess the first step is to recalculate t to be within the range of 0 to 1 again -> where 0.85 will be the new 0 and 0.3 the new 1 (or the other way around).

Then for the weight:

w = weightA * newT + weightB;

------------

here is some code i tried to isolate the problem (which is like 100% different then the actual code but might help).

Atm there is a big gap between 1.0 and 0.0 which is wrong of course cause they should be on the same position.

float[] tValuesS = {0.0, 0.2, 0.3};
float[] tValuesE = {0.85, 0.9, 1.0};
float weightStart = 15;
float weightEnd = 6;
void setup() {
size(600, 300);
textAlign(CENTER);
float tLength = 1 - tValuesE[0] + tValuesS[tValuesS.length-1];
for(int i = 0; i < tValuesE.length; i++) {
float x = map(tValuesE[i], tValuesE[0], 1+tValuesS[tValuesS.length-1], 50, width-50);
fill(i == 0 ? color(255, 0, 0) : color(0, 255, 0));
ellipse(x, 150, 10, 10);
fill(0);
text(nfc(tValuesE[i], 2), x, 180);
}
for(int i = 0; i < tValuesS.length; i++) {
float x = map(tValuesS[i]+1, 0, 1+tValuesS[tValuesS.length-1], 50, width-50);
fill(i == tValuesS.length-1 ? color(255, 0, 0) : color(0, 255, 0));
ellipse(x, 150, 10, 10);
//float w = map(tValuesS[i]+1, 0, 1, weightEnd, weightStart);
//println(w);
noFill();
ellipse(x, 150, w*10, w*10);
fill(0);
text(nfc(tValuesS[i], 2), x, 180);
}
}

Replies(4)

quarks

Re: calculate a value with a switch from 1 to 0

1 year ago

The first thing is to state that all positions ( t) along the path will be in the range >=0.0 and <1.0 which means and there is NO distance 1.0

This means that valid values of t are defined by the logical expression
(t >= 0.0 && t < 1.0)

This will simplify the algorithm dramatically.

What would be good is be a function that given the two control balls and the position of a green ball calculates its weight. This function will do that provided the ball is on the clockwise path bewteen the first and second ball, so will solve the problem as you have drawn it. (For instance what should happen to a ball at position 0.5 ?????)

/*
tFirst is distance for the first control ball
wFirst is the weight of the first control ball
tSecond is distance for the second control ball in a clockwise direction
wSecond is the weight of the second control ball in a clockwise direction
t is the distance of a ball on the clockwise path between start and end
The function will return the weight for the ball at position t
*/
float calcWeight(float tFirst, float wFirst, float tSecond, float wSecond, float t) {
float w; // the weight to calculate
if (tSecond < tFirst)
tSecond += 1.0;
if (t < tFirst)
t += 1.0;
w = ((t - tFirst)*wFirst + (tSecond - t)*wSecond) / (tSecond - tFirst);
return w;
}

BTW I have not actually tested this function you might want to try it out with some simple values to prove it works before using it with an array.

clankill3r

Re: calculate a value with a switch from 1 to 0

1 year ago

thanks so far

why < 1.0 and not <=1.0?

All the code i wrote before uses the <= 1.0.

Also i implemented your function, what do you mean by clockwise in this case?

I would assume the higher one later but that messes things up.

Now i have to pair it strange, like

the weight of 0.3 is 15 and the weight of 0.85 is 6 but i have to switch it in oder to work.

float[] tValuesS = {0.0, 0.2, 0.3};
float[] tValuesE = {0.85, 0.9, 1.0};
float weightStart = 15;
float weightEnd = 6;
void setup() {
size(600, 300);
textAlign(CENTER);
float tLength = 1 - tValuesE[0] + tValuesS[tValuesS.length-1];
for(int i = 0; i < tValuesE.length; i++) {
float x = map(tValuesE[i], tValuesE[0], 1+tValuesS[tValuesS.length-1], 50, width-50);
fill(i == 0 ? color(255, 0, 0) : color(0, 255, 0));
ellipse(x, 150, 10, 10);
float w = calcWeight(0.85, weightEnd, 0.3, weightStart, tValuesE[i]);
//float w = calcWeight(0.3, weightEnd, 0.85, weightStart, tValuesE[i]);
noFill();
ellipse(x, 150, w*10, w*10);
fill(0);
text(nfc(tValuesE[i], 2), x, 180);
}
for(int i = 0; i < tValuesS.length; i++) {
float x = map(tValuesS[i]+1, 0, 1+tValuesS[tValuesS.length-1], 50, width-50);
fill(i == tValuesS.length-1 ? color(255, 0, 0) : color(0, 255, 0));
ellipse(x, 150, 10, 10);
float w = calcWeight(0.85, weightEnd, 0.3, weightStart, tValuesS[i]);
//float w = calcWeight(0.3, weightEnd, 0.85, weightStart, tValuesS[i]);
noFill();
ellipse(x, 150, w*10, w*10);
fill(0);
text(nfc(tValuesS[i], 2), x, 180);
}
}
/*
tFirst is distance for the first control ball
wFirst is the weight of the first control ball
tSecond is distance for the second control ball in a clockwise direction
wSecond is the weight of the second control ball in a clockwise direction
t is the distance of a ball on the clockwise path between start and end
The function will return the weight for the ball at position t
*/
float calcWeight(float tFirst, float wFirst, float tSecond, float wSecond, float t) {
float w; // the weight to calculate
if (tSecond < tFirst)
tSecond += 1.0;
if (t < tFirst)
t += 1.0;
w = ((t - tFirst)*wFirst + (tSecond - t)*wSecond) / (tSecond - tFirst);
return w;
}

quarks

Re: calculate a value with a switch from 1 to 0

1 year ago

why < 1.0 and not <=1.0?

In your picture you show the path as round/ continuous so the range >=0 and <1.0 means that every point on the 'path' has a single value rather than one position having two possible values 0 or 1.

Also i implemented your function, what do you mean by clockwise in this case?

This picture might make it clearer what I mean

Now the function I provided will work for those balls at positions 0.9, 0.0 and 0.2 because they are between the first and the last in a clockwise direction.

Note that the ball at 0.5 is not between the first and the last in a clockwise position and the function will not calculate its weight correctly UNLESS you swap the actual parameters when calling the function.

So if we have
Ball            Pos       Weight
First       0.85             6
Second       0.3            15

then the weights for the three balls between first and second are calculated with

calcWeight(0.85, 6, 0.3, 15, 0.9);
calcWeight(0.85, 6, 0.3, 15, 0.0); // NOT 1.0
calcWeight(0.85, 6, 0.3, 15, 0.2);

For the ball at 0.5
calcWeight(0.3, 15, 0.85, 6, 0.5);

HTH

clankill3r

Re: calculate a value with a switch from 1 to 0

1 year ago

yes that helps.

I will implement it in the library soon and hopefully without problems.

thanks

Top Reply