Tree Formula Discussion

[witia1]

Also farming is not affecting seeds gained. With 10 farming I've gathered 30q branches and seeds from one tree.

[DatOneGuy]

Also farming is not affecting seeds gained. With 10 farming I've gathered 30q branches and seeds from one tree.

What you pull off the tree is never softcapped, we're talking about the planting of the tree.

[witia1]

But isn't it important also to gather seeds in making tree farm? Also I've typed it after reading also first page. Somehow I missed that there is more pages in thread o.O

[Saphireking65]

The amount you gather from trees is fixed, and does not change at all in any way. Different stages give different amounts, but it's always the same amount.

[Leksar]

sqrt(((qsoil+qwater+qpot+qtable)/4)*qseed) with random +-5

after each operation decimal part must be cutted, think, i posted it before

[Sitting_Bill]

^ Have you guys tested this?

BTW. How are you finding the formula? I can see there are a couple of classes of formula - ones where the components are added, and ones where they are multiplied - and then ones that combine the two, like the formula posted above . Obviously there can be different weights for all components. I don't think anything is a "soft cap" for this right? As has been written earlier in this thread, quantization errors are an issue.

Have you found the formula yet / do you have a good algorithm for searching the parameter space / shall i try and come up with one and code it up?

How much data do you have by the way? If its a lot, it might be possible to discover the formula by controlling for all but 1 input at a time.

ie just look at the data with a particular fixed SoilQ, WaterQ, TableQ, PotQ - and plot TreeQ versus SeedQ. I doubt you have this much data however, 6 dimensional space is hard to scan :S, and I guess noone can be bothered to do controlled treefarming tests purely for this test...

I'm just about to start up mine, I'll be sure to upload my data when I do

[Leksar]

^ Have you guys tested this?

BTW. How are you finding the formula? I can see there are a couple of classes of formula - ones where the components are added, and ones where they are multiplied - and then ones that combine the two, like the formula posted above . Obviously there can be different weights for all components. I don't think anything is a "soft cap" for this right? As has been written earlier in this thread, quantization errors are an issue.

Have you found the formula yet / do you have a good algorithm for searching the parameter space / shall i try and come up with one and code it up?

How much data do you have by the way? If its a lot, it might be possible to discover the formula by controlling for all but 1 input at a time.

ie just look at the data with a particular fixed SoilQ, WaterQ, TableQ, PotQ - and plot TreeQ versus SeedQ. I doubt you have this much data however, 6 dimensional space is hard to scan :S, and I guess noone can be bothered to do controlled treefarming tests purely for this test...

I'm just about to start up mine, I'll be sure to upload my data when I do

it's not too hard, if you have some data and made some experiments. Me and my friend Gray Rat was interested in formula in w3, so his friend helped us with experiments and Koya helped me to find right direction

[burgingham]

Once again alpha made himself the laughing stock of the forums. Not that I am a big fan of Leksar, but thanks for this one

[DatOneGuy]

Will be testing that, we still don't have a lot of data. Maybe 20 trees here. I don't trust my data from W3 so I haven't added it. I've been trying to obtain some nice same Q stuff so I can test things one at a time, but that would start with a few tree pots of the same exact Q, and I've had problems there.

[Leksar]

Once again alpha made himself the laughing stock of the forums. Not that I am a big fan of Leksar, but thanks for this one

as i said before, alpha is just a russian bot who says nonsense on all sides, so i can't understant, why do u react to him