def point_poisson():
global occupation_poisson
while len(active) > 0:
check = np.random.randint(0, len(active))
found = None
for n in range(k):
s = _random_vector_function.random_vector_function(
r, math.floor(r * 2))
sample = []
for i in range(0, len(s)):
sample.append(s[i] + active[check][i])
gcol = math.floor(sample[0] / w)
grow = math.floor(sample[1] / w)
ggrid = gcol + grow * cols
if gcol > -1 and grow > -1 and gcol < cols and grow < rows:
ok = True
#check distance between selected point and points in the around grids
nums = [-1, 0, 1]
num1s = [-1, 0, 1]
for num in nums:
for num1 in num1s:
if 0 <= gcol + num1 < cols and 0 <= grow + num < rows:
index = (gcol + num1) + (grow + num) * cols
if grid[index] is None:
pass
else:
d = _distance2d.dist(grid[index][0],
grid[index][1], sample[0],
sample[1])
if d < r:
ok = None
if ok is True:
found = True
grid[gcol + grow * cols] = sample
active.append(sample)
break
if found is None:
del active[check:(check + 1)]
for i in range(len(grid)):
if grid[i] is None:
pass
else:
if (grid[i][0] < width - roundRadius[0]) and (
grid[i][0] > roundRadius[0]) and (
grid[i][1] < height - roundRadius[0]) and (
grid[i][1] > roundRadius[0]):
Circles.append(
_Circle.Circle(grid[i][0], grid[i][1], roundRadius[0],
width, height))
occupation_poisson = occupation_poisson + 1
# enlarge the particles
_generateParticle.generateCircles_p(ranges[0], (maximums[0] - 2), Circles)
print('poisson disk sampling', occupation_poisson)
def generateCircles_p(mindis, maxi, Circles):
growing = []
for k in range(len(Circles)):
if Circles[k].g is True:
growing.append([Circles[k].x, Circles[k].y])
while len(growing) > 0:
for i in range(len(Circles)):
if Circles[i].g is True:
if Circles[i].r > maxi:
Circles[i].g = None
growing.remove([Circles[i].x, Circles[i].y])
break
elif Circles[i].edge() is None:
growing.remove([Circles[i].x, Circles[i].y])
break
else:
for j in range(len(Circles)):
if (j != i) and (
(Circles[i].x - mindis) < Circles[j].x <
(Circles[i].x + mindis)) and (
(Circles[i].y - mindis) < Circles[j].y <
(Circles[i].y + mindis)):
dis = _distance2d.dist(Circles[i].x, Circles[i].y,
Circles[j].x, Circles[j].y)
if dis < (Circles[i].r + Circles[j].r):
Circles[i].g = None
growing.remove([Circles[i].x, Circles[i].y])
break
Circles[i].grow()
def radii(q, initial_radius, range1, maximum1, Circles, cols1, w1, width,
height):
valid = True
yaxis = math.floor(q / cols1)
xaxis = math.floor(q - yaxis * cols1)
w_use = w1
x = np.random.randint(math.floor(xaxis * w_use),
math.floor((xaxis + 1) * w_use))
y = np.random.randint(math.floor(yaxis * w_use),
math.floor((yaxis + 1) * w_use))
#if the randomly selected point in the void grid is too close to borders, it will be deleted.
m = initial_radius
min_boundary = [x, y, (width - x), (height - y)]
min_b = np.amin(min_boundary)
if min_b < m:
valid = None
#in the reseaching sphere area with 'mindis' radius
min_around = []
if valid is True:
for j in range(len(Circles)):
if ((x - range1) < Circles[j].x <
(x + range1)) or ((y - range1) < Circles[j].y < (y + range1)):
k = _distance2d.dist(x, y, Circles[j].x,
Circles[j].y) - Circles[j].r
if (k < m):
valid = None
break
else:
min_around.append(k)
#generate particles
if valid is True:
if not min_around:
min_a = 0
else:
min_a = np.amin(min_around)
if min_a <= min_b:
if min_a <= maximum1:
Circles.append(_Circle.Circle(x, y, min_a, width, height))
return min_a
else:
Circles.append(_Circle.Circle(x, y, maximum1, width, height))
return maximum1
elif min_a > min_b:
if min_b <= maximum1:
Circles.append(_Circle.Circle(x, y, min_b, width, height))
return min_b
else:
Circles.append(_Circle.Circle(x, y, maximum1, width, height))
return maximum1
else:
return False
