@DanielK1 I’m switching our discussion from yesterday to here, it’s your turn now

Hi,
I was wondering if there is a better way to divide a curve in smaller segments with ever increasing steps, so that the last element has a specific length (e.g. 5mm) here is my approach: Sadly after remapping the last element has a slightly bigger value than optimal (<5mm). Also it only works with “symetric” graphs-mappers…

maybe someone knows this:
how can one “blend” between those 2 point distributions?
It would be nice if it would behave more like A at the start of the curve and more like B at the end.
there are t parameters for both.

thanks! sadly no – what I am looking for is that the last segment has a specific length (e.g 5mm) and gets smoothly and continuously smaller until it gets to the start of the curve.

You want series where the first number is 5 (5, 4.8, 4.4 , 4.1 … etc) and the sum equal to the length of the curve
i think this is not possible, maybe you can use Genome to solve that and you will not get precise result;
and the series based on what ?

@DanielK1
thanks, I still don’t quite understand how you do it and what I am doing wrong since it doesn’t produce the linesegment with the right lenght.see here:

smooth-start-v8.gh (20.4 KB)
I can’t set the nummer of items though, we will see if it makes any difference – I allway can make the curve longer if the acceleration is to high ( it is a speedramp for a robot )

I thought that applying a scale factor geometrically to the initial segment would give you a series of n gradually decreasing pieces. The problem is to determine that factor so that the lengths all add up to the length of the curve. For this (with the help of @dharman,Newton's Method - #3 by akilli ) I used Newton’s Method in a Python component to come up with a number for the scale between 0 and 1 (though, its usually pretty close to 1).

really impressive! How do you know those things? – very interesting. I will look into the script and newtons method - Maybe I can modify it so the length distribution has a “s-curve”…

(gh love)

btw: his is what the script does (regulating speed):

Hi Konrad, the problem is that Evaluate and Parameter of curves not reliable, due to the unequal positions of control points. This property makes it inaccurate.

I was inspired by @akilli to use an actual length and remake formula. It seems to me that it’s working quite OK. Let’s have a look.

I created domains from a combination of “the desired length” and “lengths of equally distributed points”. I then selected values from 0 to 1 from each domain to create a smooth transition.

You can as well control the curvature of transition.

Edit: I realized that actually, you need to distribute the points and then create polyline which is going inaccuraty in the case of a lot of curved curves.