R4tt3xx
I want to believe
- Joined
- Apr 10, 2005
- Posts
- 2,176
- Location
- South Africa
- Society
- Freelancer
- Avatar Name
- Alexis Sky Greenstar
Warning the attached file does use macros, please look through the code or use in a virtual PC to ensure your safety.
The file https://drive.google.com/open?id=0Bz393x-q88QWdmtZX0J6cHczS0k is a little experiment I am currently running but can be adapted.
Basically what it does is generate a set of points laid out as a fermat spiral using 0.618 as its turn angle, the circle is then transformed into a square the same size as a typical mining radius (108m). By playing around the sheet can be adjusted to create any number of points and any size array. The 2 numbers under x and y in the centroid block is the middle of the array.
The main purpose of it's creation was to look at creating a model for mining blah blah blah..
I am using it in it's current form to determine where the most optimal location would be to place a charge and cover the most likely locations of claims in the area, which is the centroid, ie average, ie where all the points connect.
Imagine dropping probes in a set area and the value of those probes ie 0.05 ped each, settle in different locations inside this mining square, this mining square is all there is, a repeating area where everyone's mining spent probes are pooled together, If one distributes 30 probes, only a max of 20 will be found at any one time based on this sheet.
But it gets worse... It is possible to slightly tweak the rules so that there is no way of reclaiming spent probes, by making the grid slightly smaller, so that rounding errors occur, one can hide the spent probes on the boundaries of the square, delete cell c4 on the control sheet to see this happen. The problem is that one does not know what configuration the distribution is as there can be other ways to distribute the spent probes so that they are not easily discovered.
The most powerful counter to this is the centroid, if one uses the centroid in this alternate layout, the hitrate jumps from 0 % to 2/30 ie 6.667 %, One could just simple blow up the edges but you run the risk of loss by not covering the centre.
This is accomplished by changing the base of the distance, the log option on the control sheet. By changing the log to its smallest +- 0.01, 96.6667% of spent probes can be recovered.
This sheet has a long ways to go, but it can I think become a power house for modeling.
What I would like to do and I cant find a quick way of doing it, is to work out all the centroid coordinates for the base variations, and then centroid that in order to produce an average over all the averages.
The most beautiful thing if such a model is being used, is how fair it would be, all spent probes would sit in the array, waiting for recovery.
Again, please comment..
Later
The file https://drive.google.com/open?id=0Bz393x-q88QWdmtZX0J6cHczS0k is a little experiment I am currently running but can be adapted.
Basically what it does is generate a set of points laid out as a fermat spiral using 0.618 as its turn angle, the circle is then transformed into a square the same size as a typical mining radius (108m). By playing around the sheet can be adjusted to create any number of points and any size array. The 2 numbers under x and y in the centroid block is the middle of the array.
The main purpose of it's creation was to look at creating a model for mining blah blah blah..
I am using it in it's current form to determine where the most optimal location would be to place a charge and cover the most likely locations of claims in the area, which is the centroid, ie average, ie where all the points connect.
Imagine dropping probes in a set area and the value of those probes ie 0.05 ped each, settle in different locations inside this mining square, this mining square is all there is, a repeating area where everyone's mining spent probes are pooled together, If one distributes 30 probes, only a max of 20 will be found at any one time based on this sheet.
But it gets worse... It is possible to slightly tweak the rules so that there is no way of reclaiming spent probes, by making the grid slightly smaller, so that rounding errors occur, one can hide the spent probes on the boundaries of the square, delete cell c4 on the control sheet to see this happen. The problem is that one does not know what configuration the distribution is as there can be other ways to distribute the spent probes so that they are not easily discovered.
The most powerful counter to this is the centroid, if one uses the centroid in this alternate layout, the hitrate jumps from 0 % to 2/30 ie 6.667 %, One could just simple blow up the edges but you run the risk of loss by not covering the centre.
This is accomplished by changing the base of the distance, the log option on the control sheet. By changing the log to its smallest +- 0.01, 96.6667% of spent probes can be recovered.
This sheet has a long ways to go, but it can I think become a power house for modeling.
What I would like to do and I cant find a quick way of doing it, is to work out all the centroid coordinates for the base variations, and then centroid that in order to produce an average over all the averages.
The most beautiful thing if such a model is being used, is how fair it would be, all spent probes would sit in the array, waiting for recovery.
Again, please comment..
Later