Code center > Bugs and fixes

Cost Multiplier changing

<< < (2/2)

EricL:
Well, your target population is 500 and with a lower range of 1% and dynamic costs enabled, the code is correctly lowering CostX as the hetertropic population is below 495.  You only have 251 heterotrofs in your simulation.  It makes sense that CostX will decrease over time until it hits 0....  Frankly, I don't see a bug...

Testlund:
The problem is that the Cost Multiplier setting doesn't stay put. If you set it to 1, click ok, then go back it will display another value. Even if I turn off dynamic costs and set the other two override costs to 0 the Cost multiplier will change.

EricL:
If Dynamic Costs are enabled, the cost multiplier will not stay put.  This is by design and what Dynamic Costs are all about - changing the cost muliplier automatically.  You can set the current value manually and it will start automatically adjusting it from that new value, but as long as dynamic costs are enabled, CostX will get modified dynamically if the population is outside the range.

When Dynamic Costs are disabled, CostX should not change.  I can't seem to make it happen, but perhaps because you are setting the reinstatement level to 0, the code is attempting to reinstate a previous CostX since the population is now above the threshold of 0.  Set it to somethign big and see if CostX still changes on you when dynamic costs are disabled.

On a related note, the reason the dynamic costs population only refers to heterotrofs is that veggies with reproduction genes can result in sudden swings in autotropic population numbers in a single cycle (when the veggy population first falls below the max veggy number, they all reproduce).  Now that we have the ability to disable reproduction for specific species, we could change it so that the population number refers to all bots in the sim if there was interest...

Testlund:
Ok, I've been fiddleing with the cost settings to try and understand what's going on. It looks like the value for the Cost Multiplier changes depending on the amount of bots that exists in the sim compared to the cost values. Maybe that's what it's supposed to do but I might have missunderstood it.
But it still appears that this function doesn't update correctly if you for instance chose to disable dynamic costs. Like I said the cost multiplier changes sometimes to another value just by leaving and entering the tab.

Testlund:

--- Quote from: EricL ---...we could change it so that the population number refers to all bots in the sim if there was interest...
--- End quote ---

Yes, I've been thinking about that too. It whould be very useful for my evosim which are only autotrophs!   Maybe just putting a checkbox in that says "Include autotrophs" or something.

Also, I find this dynamic cost thing difficult to understand. Thought I got a hang of it awhile ago when I was playing around with it, but I think I forgot. For instance the lower range value didn't make any sense to me how it worked. Ok, I need to play with this some more.  

Navigation

[0] Message Index

[*] Previous page