Currently Online

Latest Posts

Topic: Do trees grow on mountains?

einstein13
Avatar
Joined: 2013-07-29, 00:01
Posts: 1118
Ranking
One Elder of Players
Location: Poland
Posted at: 2015-03-16, 01:21

Ok, I've managed to do the test for foresters - woodcutters. It is useful for players face-wink.png miu factors defined as "the possibility of full grow of tree planted by the forester" mountains/ not buildable grounds were filled with barren steppe (summer) in small spots. 5 hours test shown that:

Less than 6%:

  • Desert mountains
  • Summer barren steppe
  • Winter mountains
  • Summer mountains
  • Deserts 2 - 4, form desert
  • Summer Beach

10..30%

  • Winter ice
  • Desert 1 from desert
  • Wasteland & Winter Beach

35%..60%

  • Wasteland Mountains
  • Winter Snow
  • Beach, Dry soil, High mountain meadow from Desert
  • Ashes 2, Igneous rocks from Wasteland
  • Winter taiga
  • Summer mountain meadow

60..80%

  • Summer & desert steppe
  • Wasteland hard grounds
  • Wasteland ashes 1
  • Winter tundra (4 & 1)
  • Desert mountain meadow
  • Summer meadow 1

85..100%

  • Winter tundras 2 & 3
  • Summer meadows 2-4
  • Desert meadow

Quotes:

(...) with a mix of all four mountain types, not even one.

I will never try mixing the terrain types. Why? Because then I have to check

  • 50*49/2=1225 types of mixes for just 2 terrain types
  • 19600 for 3 terrain types
  • ... even more for more terrain type mix

It is TOO MUCH for me face-smile.png 50 types of terrain was a long way (about 12 hours of work)

So, I'd say the chance of a tree growing on a mountain terrain of any kind could be around 3%

For desert is 0.0-0.2%, but for wasteland mountains it is between 37.5-43.5%. Your test isn't telling about any specific kind of mountains, but about the mix.


einstein13
calculations & maps packages: http://wuatek.no-ip.org/~rak/widelands/
backup website files: http://kartezjusz.ddns.net/upload/widelands/

Top Quote
kaputtnik
Avatar
Joined: 2013-02-18, 20:48
Posts: 2397
OS: Archlinux
Version: current master
Ranking
One Elder of Players
Location: Germany
Posted at: 2015-03-16, 09:21

I am no scientist, so i do not fully understand what you are talking about face-upset.png

(...) with a mix of all four mountain types, not even one.

I will never try mixing the terrain types. Why? Because then I have to check

I don't understand why you have to check this. Isn't it possible to make a specific tree do grow on all mountain terrains with the terrain affinities in an easy way?

I think we have a conflict here, because we have the old image of four different worlds with four different terrain types with four different tree species. Bringing back the old differences of each world into oneworld is imho partly wrong.

I think we have to break with the old image of four different worlds with four different terrain types.

Instead we should think about f.e.:

  • let only coniferes grow on mountains
  • let umbrella trees do grow only on dark terrains types
  • let palm trees do grow only on yellowish terrains
  • and so on

Lets make rules for treegrowing which depends on other things than the old four world image.

Just from a non scientist face-smile.png


Fight simulator for Widelands:
https://wide-fighter.netlify.app/

Top Quote
einstein13
Avatar
Joined: 2013-07-29, 00:01
Posts: 1118
Ranking
One Elder of Players
Location: Poland
Posted at: 2015-03-16, 16:10

kaputtnik wrote:

I am no scientist, so i do not fully understand what you are talking about face-upset.png

(...) with a mix of all four mountain types, not even one.

I will never try mixing the terrain types. Why? Because then I have to check

I don't understand why you have to check this.

If you want to check the behaviour of 1 type of terrain and you have n terrains you have to check n different cases to have all the data. (In our situation we have n=50)

Ok, let try 2 different types of terrain at once. What will it be? type 1 with type 2; type 1 with type 3; ... , type n-1 with type n. That is n*(n-1)/2 cases. For every n types you have n-1 different types. But if you simply multiply n and n-1 you will have "type 1 with type 2" and "type 2 with type 1" which is counted twice. All those situations are counted twice, so we divide by 2. This math is from basic school math face-wink.png

With 3 different types of terrain you have more complex situation, but the equation is very simmilar: n(n-1)(n-2)/6. why 6? for (1,2,3) you have 6 combinations that will give the same result for us: (1,2,3), (1,3,2), (2,1,3), (2,3,1), (3,1,2), (3,2,1). So we have to divide by 6.

For "k" different types of terrain types mixed the equation:

n!/(k!*(n-k)!)

This is binomial coefficient (wikipedia) and that is a bit higher math face-wink.png Ohh... "!" is a factorial (wikipedia)

Isn't it possible to make a specific tree do grow on all mountain terrains with the terrain affinities in an easy way?

I don't think so. Not in current model. You suggested to make additional models for map editor- this is possible, but not very easy :).

I think we have a conflict here, (...)

Where is the conflict? I didn't make any suggestions! I've made pure experiment. No opinion included. face-wink.png

the old image of four different worlds with four different terrain types with four different tree species.

I don't think that is a bad idea to keep the image. For me it is a beauty of the game face-smile.png I can easily make whole world ecosystem with those parameters.

Instead we should think about f.e.:

  • let only coniferes grow on mountains
  • let umbrella trees do grow only on dark terrains types
  • let palm trees do grow only on yellowish terrains
  • and so on

It is almost the same. I didn't test trees-terrain at all. I will do it soon, but I don't have much time. For me it was very hard to make the previous test.

Lets make rules for treegrowing which depends on other things than the old four world image.

It is now.


einstein13
calculations & maps packages: http://wuatek.no-ip.org/~rak/widelands/
backup website files: http://kartezjusz.ddns.net/upload/widelands/

Top Quote
kaputtnik
Avatar
Joined: 2013-02-18, 20:48
Posts: 2397
OS: Archlinux
Version: current master
Ranking
One Elder of Players
Location: Germany
Posted at: 2015-03-16, 18:08

Hmmm... i don't understand it anyway. Sorry, but i think a code where the results could only be retrieved with a huge number of tests seems to me a bit unpredictable....

But my thoughts are surely a result of not understanding. I have to trust you and have to be silent face-smile.png


Fight simulator for Widelands:
https://wide-fighter.netlify.app/

Top Quote
king_of_nowhere
Avatar
Topic Opener
Joined: 2014-09-15, 18:35
Posts: 1668
Ranking
One Elder of Players
Posted at: 2015-03-16, 19:10

I will never try mixing the terrain types. Why? Because then I have to check

50*49/2=1225 types of mixes for just 2 terrain types 19600 for 3 terrain types ... even more for more terrain type mix

I didn't mix the mountain types because I wanted to check on the boundaries between them. I simply did it to save time: after trying mountain 1, and seeing trees did not grow on them, I guessed no trees would grow on any summer mountain terrain, so I used all of them together in the next test. If no tree grew on it (as it happened), it meant that all types of mountains are sterile, and I don't need to check further. If, instead, trees are produced, then I would have to make further inquiries and see which mountain type supported the trees. Basically, I saved two tests. I haven't checked the mmountains of the other kind of terrains yet.

The only boundary I want to check is the one with water, because it is quite recurrent in many maps. And then, only with meadows. If I see that trees grow well on a water-meadow boundary, then I will assume that water does not significnatly hamper tree growth. I do not have your patience to try all the terrains.


Top Quote
einstein13
Avatar
Joined: 2013-07-29, 00:01
Posts: 1118
Ranking
One Elder of Players
Location: Poland
Posted at: 2015-03-16, 21:42

kaputtnik wrote:

Hmmm... i don't understand it anyway. (...) I have to trust you and have to be silent face-smile.png

It is good that you tell me that you don't understand. I will ask you to read a text summarizing all the work we will do AND to check if everything is clear with tables we will create (the same will be for king_of_nowhere). I can explain everything from very basic things, but here it is no place for that face-smile.png Now you can trust me and argue if I'm wrong in the future face-wink.png

king_of_nowhere wrote:

I didn't mix the mountain types because I wanted to check on the boundaries between them. (...) If no tree grew on it (as it happened) (...)

From my tests there is a difference only for one type of mountains: wasteland mountains. They are pretty fertile. Please check it with your type of test. If you can check it with: one type of mountain in one test (4 times) and mix of all types of mountains (1 test). I'm curious if my values are different than your AND if mix of mountains make any difference.

The only boundary I want to check is the one with water, because it is quite recurrent in many maps. (...) then I will assume that water does not significnatly hamper tree growth

For sure this can make huge differences! water should have set of [temperature: ..., fertility: ..., humidity: 1]. That can make the desert very fertile for another kind of tree face-smile.png For example for pines, birch or other strange trees face-wink.png

Meadows are not so frequent. I find the sand (beach) more frequent as a neighbour to water. Consider this type of terrain too face-smile.png

I do not have your patience to try all the terrains.

This is not a patience, this is motivation. I did the work not for me, but for you. Another man is kaputtnik- he will be the next. But this work will last longer. I need help with that. For sure. Because counting the trees is quite... time spending and boring face-tongue.png

And I've found that current saving/ opening map model is not so good. I can have 50 maps named "no name" when choosing the map for test. It is quite annoying. If I forget about renaming the map I will have lots of troubles to be sure which kind of tree I'm currently testing. This is annoying... face-confused.png


einstein13
calculations & maps packages: http://wuatek.no-ip.org/~rak/widelands/
backup website files: http://kartezjusz.ddns.net/upload/widelands/

Top Quote
GunChleoc
Avatar
Joined: 2013-10-07, 15:56
Posts: 3324
Ranking
One Elder of Players
Location: RenderedRect
Posted at: 2015-03-17, 14:15

einstein13 wrote:

And I've found that current saving/ opening map model is not so good. I can have 50 maps named "no name" when choosing the map for test. It is quite annoying. If I forget about renaming the map I will have lots of troubles to be sure which kind of tree I'm currently testing. This is annoying... face-confused.png

I am already working on that - you are welcome to add your thoughts to the thread: https://wl.widelands.org/forum/topic/1676/


Busy indexing nil values

Top Quote
einstein13
Avatar
Joined: 2013-07-29, 00:01
Posts: 1118
Ranking
One Elder of Players
Location: Poland
Posted at: 2015-03-17, 15:40

GunChleoc wrote:

einstein13 wrote:

And I've found that current saving/ opening map model is not so good. (...)

I am already working on that - you are welcome to add your thoughts to the thread: https://wl.widelands.org/forum/topic/1676/

Yes, I noticed your work and I will be verry happy to see the results! face-smile.png I also wrote some words on that topic face-smile.png Thank you for your work! face-smile.png


einstein13
calculations & maps packages: http://wuatek.no-ip.org/~rak/widelands/
backup website files: http://kartezjusz.ddns.net/upload/widelands/

Top Quote