Fractal terrian program making 200-year old spike pits.

Author

Message

Narf The Mouse

20

Years of Service

User Offline

Joined: 16th Jul 2004

Location:

Posted: 16th Jul 2004 12:32

Link

I set every point on the heightmap to -1 to differenciate between points set and points not set. then I put random heights on the heightmap from 0,0 to currently 32,32 using a for/next loop and stepping by half the map size, so 9 height points. The program then is supposed to step accross the map at half intervals, check to see if the current point is at -1, if it is, take the average of all points at the current step distance x,y or both away that aren't at -1, then input the average into it's current square. Once it completes a map, it divides both distances it's stepping by by /2 and goes again until it's done the map at 1,1 intervals.

So, can someone take a look at the code and tell me what I'm doing wrong, please?

I've been at this for a while.

Cheese!

Back to top

Profile PM

indi

22

Years of Service

User Offline

Joined: 26th Aug 2002

Location: Earth, Brisbane, Australia

Posted: 16th Jul 2004 13:40

Link

I was given this function a while back which smooths a matrix.
the paramters of the function show what to do in order to use it.

+ Code Snippet

function smooth_matrix(matnum,tilex,tilez,times)
for multiple=1 to times

for x=1 to tilex-1
for z=1 to tilez-1

a=get matrix height(matnum,x-1,z+1)
b=get matrix height(matnum,x,z+1)
c=get matrix height(matnum,x+1,z+1)
d=get matrix height(matnum,x+1,z)
e=get matrix height(matnum,x+1,z-1)
f=get matrix height(matnum,x,z-1)
g=get matrix height(matnum,x-1,z-1)
h=get matrix height(matnum,x-1,z)
total=a+b+c+d+e+f+g+h
av#=total/8
set matrix height matnum,x,z,av#
next z
next x
next multiple

for x=0 to tilex
set matrix height matnum,x,0,get matrix height(matnum,x,1)
next x

for x=0 to tilex
set matrix height matnum,x,tilez,get matrix height(matnum,x,tilez-1)
next x

for z=0 to tilez
set matrix height matnum,0,z,get matrix height(matnum,1,z)
next z

for z=0 to tilez
set matrix height matnum,tilex,z,get matrix height(matnum,tilex-1,z)
next z

update matrix matnum
endfunction

If no-one gives your an answer to a question you have asked, consider:- Is your question clear.- Did you ask nicely.- Are you showing any effort to solve the problem yourself

Back to top

Profile PM Email

Phaelax

DBPro Master

21

Years of Service

User Offline

Joined: 16th Apr 2003

Location: Metropia

Posted: 16th Jul 2004 17:49

Link

I have no idea what you're talking about. But if smoothing the matrix out is what you want, then here's a routine I made just for that.

_smooth_matrix:

dim heights(tx+1,tz+1)

for z = 0 to tz
      for x = 0 to tx
         heights#(x,z) = get matrix height(1,x,z)
      next x
   next z

for z = 0 to tz
      for x = 0 to tx
         count = 0
         h1# = 0
         h2# = 0
         h3# = 0
         h4# = 0
         h5# = 0
         h6# = 0
         h7# = 0
         h8# = 0
         if z < tz
            if x > 0 then h1# = heights#(x-1,z+1) : inc count
            h2# = heights#(x,z+1) : inc count
            if x < tx then h3# = heights#(x+1,z+1) : inc count
         endif

if x > 0 then h4# = heights#(x-1,z) : inc count
         if x < tx then h5# = heights#(x+1,z) : inc count

if z > 0
            if x > 0 then h6# = heights#(x-1,z-1) : inc count
            h7# = heights#(x,z-1) : inc count
            if x < tx then h8# = heights#(x+1,z-1) : inc count
         endif

avg# = (h1#+h2#+h3#+h4#+h5#+h6#+h7#+h8#) / count

set matrix height 1,x,z,avg#
      next x
   next z

update matrix 1

undim heights()

RETURN

+ Code Snippet

_smooth_matrix:

   dim heights(tx+1,tz+1)

   for z = 0 to tz
      for x = 0 to tx
         heights#(x,z) = get matrix height(1,x,z)
      next x
   next z

   for z = 0 to tz
      for x = 0 to tx
         count = 0
         h1# = 0
         h2# = 0
         h3# = 0
         h4# = 0
         h5# = 0
         h6# = 0
         h7# = 0
         h8# = 0
         if z < tz
            if x > 0 then h1# = heights#(x-1,z+1) : inc count
            h2# = heights#(x,z+1) : inc count
            if x < tx then h3# = heights#(x+1,z+1) : inc count
         endif

         if x > 0 then h4# = heights#(x-1,z) : inc count
         if x < tx then h5# = heights#(x+1,z) : inc count

         if z > 0
            if x > 0 then h6# = heights#(x-1,z-1) : inc count
            h7# = heights#(x,z-1) : inc count
            if x < tx then h8# = heights#(x+1,z-1) : inc count
         endif

         avg# = (h1#+h2#+h3#+h4#+h5#+h6#+h7#+h8#) / count

         set matrix height 1,x,z,avg#
      next x
   next z

   update matrix 1

   undim heights()

RETURN

The problem with indi's is that its not averaged out properly. After a point's new averaged height is configured, it's written immediately back to the height data. More points will use that when their heights are averaged. Therefore, each point ends up being averaged with not the original heights, but with averaged height values. To fix this, you have an array of all original heights. Each new averaged height is configured from that array, and the result can be written to the matrix structure without worry of overwriting needed data. Also, his routine doesn't affect edge tiles, but that's probably just a preference.(which also minimizes code)

"eureka" - Archimedes

Back to top

Profile PM Email Website

Narf The Mouse

20

Years of Service

User Offline

Joined: 16th Jul 2004

Location:

Posted: 25th Jul 2004 12:36

Link

Thanks. I had an allergy attack and kept forgetting about this. That does solve the spikes, but I would like to know why there are spikes.

Cheese!

Back to top

Profile PM

Sorry your browser is not supported!

DarkBASIC Professional Discussion / Fractal terrian program making 200-year old spike pits.