Questions 

I have noticed that movement in a circle is measured as distance. How about
a new measure; distance from the point of origin?

Also, has anyone noticed that the ground level is a little high (the sticks
seem to sink into the ground).

Forums: 
Maciej Komosinski's picture

> I have noticed that movement in a circle is measured as distance. How about
> a new measure; distance from the point of origin?

You can adjust precision of distance measurements, in
properties of each group.

> Also, has anyone noticed that the ground level is a little high (the sticks
> seem to sink into the ground).

Because they have some mass, and the ground is
not perfectly firm.

Had the same question.
This could maybe be done by modifying the onStep method in the .expdef file
you are using, to override one of the user parameters.
NOT tried, though
succes!
Frans

"Maciej Komosinski" schreef in bericht
news:3DF87C9F.1090500@cs.put.poznan.pl...
> > I have noticed that movement in a circle is measured as distance. How
about
> > a new measure; distance from the point of origin?
>
> You can adjust precision of distance measurements, in
> properties of each group.
>
> > Also, has anyone noticed that the ground level is a little high (the
sticks
> > seem to sink into the ground).
>
> Because they have some mass, and the ground is
> not perfectly firm.
>

To evaluate your distance every step (instead of after a life-time) adjust
the parameters in:
Experiment -> Populations -> Creatures -> Performance Sampling Value (PSV)

If the PSV is set to 1, the fitness (e.g. distance) is measured every single
step. Then, the total distance of the trajectory from the point of origin
can be the fitness value of an individual.
Note that also a shaking Fram will result in an increased distance value
(because every step there is some little change in distance) although the
total distance from the point of origin will be near zero.
If the PSV is set to the total lifetime (for example 1000) the
fitness/distance is measured as the distance between the point of origin and
the point of death as a straight line (not taking into account the
trajectory).
Some experimenting is thus required to set the PSV appropriate value

Hope this is what you wanted to know. Success!

Walter

Walter,

Your solution would give erratic results. I observed that many creatures run
(roughly) in circles. Thus one running a long distance, almost making the
full circle, would be recorded as having travelled only a short disrance. A
more lazy crature running only say one six-th of the same circle would be
rated far better, whilst it has reached only half the distance from its
place of birth.
What I was looking for (I can not speak for Paul) is storing a value (-1)
upon birth and with every step (10 steps?) check the position from place of
birth. If the current distance exceeds the stored value, update the stored
value with the current distance.
Hope this clarifies. I got no time to work it out now. Maybe during
Christmas break.

Frans Verbaas

"Walter de Back" schreef in bericht
news:atfpf9$8i1$1@cancer.cs.put.poznan.pl...
> To evaluate your distance every step (instead of after a life-time) adjust
> the parameters in:
> Experiment -> Populations -> Creatures -> Performance Sampling Value (PSV)
>
> If the PSV is set to 1, the fitness (e.g. distance) is measured every
single
> step. Then, the total distance of the trajectory from the point of origin
> can be the fitness value of an individual.
> Note that also a shaking Fram will result in an increased distance value
> (because every step there is some little change in distance) although the
> total distance from the point of origin will be near zero.
> If the PSV is set to the total lifetime (for example 1000) the
> fitness/distance is measured as the distance between the point of origin
and
> the point of death as a straight line (not taking into account the
> trajectory).
> Some experimenting is thus required to set the PSV appropriate value
>
> Hope this is what you wanted to know. Success!
>
> Walter
>
>