-
Notifications
You must be signed in to change notification settings - Fork 7
/
Copy pathFractalGeneratorV3.py
286 lines (240 loc) · 10.9 KB
/
FractalGeneratorV3.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
"""
@author: DJZ-Nodes
Fractal Generator V3 - A CUDA-accelerated node that generates fractal art using various fractal types with advanced controls
"""
import numpy as np
from PIL import Image
import torch
import colorsys
import math
class FractalGeneratorV3:
"""A ComfyUI node that generates fractal art with CUDA acceleration"""
@classmethod
def INPUT_TYPES(s):
return {
"required": {
"width": ("INT", {"default": 1024, "min": 64, "max": 4096}),
"height": ("INT", {"default": 1024, "min": 64, "max": 4096}),
"max_iterations": ("INT", {"default": 500, "min": 50, "max": 2000}),
"preset": (["Custom", "Classic Mandelbrot", "Julia Set", "Burning Ship", "Tricorn", "Newton"], {"default": "Classic Mandelbrot"}),
"zoom_level": ("FLOAT", {"default": 1.0, "min": 0.1, "max": 100.0, "step": 0.1}),
"color_preset": ([
"Classic White-Grey",
"Electric Blue",
"Fire",
"Rainbow",
"Deep Space",
"Ocean",
"Forest",
"Psychedelic"
], {"default": "Classic White-Grey"}),
"power": ("FLOAT", {"default": 2.0, "min": 2.0, "max": 5.0, "step": 0.1}),
"escape_radius": ("FLOAT", {"default": 2.0, "min": 1.0, "max": 10.0, "step": 0.1}),
"color_cycles": ("FLOAT", {"default": 1.0, "min": 0.1, "max": 10.0, "step": 0.1}),
"smooth_coloring": ("BOOLEAN", {"default": True}),
},
"optional": {
"x_center": ("FLOAT", {"default": -0.75, "min": -2.0, "max": 2.0, "step": 0.0001}),
"y_center": ("FLOAT", {"default": 0.0, "min": -2.0, "max": 2.0, "step": 0.0001}),
}
}
RETURN_TYPES = ("IMAGE",)
FUNCTION = "generate_fractal"
CATEGORY = "DJZ-Nodes"
def get_preset_coordinates(self, preset, zoom_level):
presets = {
"Classic Mandelbrot": (-0.75, 0.0, 1.0),
"Julia Set": (0.0, 0.0, 1.5),
"Burning Ship": (-0.5, -0.5, 0.8),
"Tricorn": (0.0, 0.0, 1.2),
"Newton": (0.0, 0.0, 1.0),
}
if preset in presets:
x, y, base_zoom = presets[preset]
actual_zoom = base_zoom * zoom_level
return x, y, actual_zoom
return None
def get_color_function(self, preset):
def classic_white_grey(value):
grey = int(255 * (1 - value))
return (grey, grey, grey)
def electric_blue(value):
if value > 0.95:
return (255, 255, 255)
blue = int(255 * (0.5 + 0.5 * value))
green = int(128 * value)
return (int(60 * value), green, blue)
def fire(value):
if value < 0.33:
return (int(255 * (0.5 + 1.5 * value)), 0, 0)
elif value < 0.66:
v = (value - 0.33) * 3
return (255, int(255 * v), 0)
else:
v = (value - 0.66) * 3
return (255, 255, int(255 * v))
def rainbow(value):
hue = value % 1.0
sat = 0.9
val = 0.9
rgb = colorsys.hsv_to_rgb(hue, sat, val)
return tuple(int(255 * x) for x in rgb)
def deep_space(value):
if value > 0.95:
return (255, 255, 255)
elif value > 0.90:
star_bright = int(200 * (value - 0.90) * 10)
return (star_bright, star_bright, star_bright)
hue = 0.75 + value * 0.15
sat = 0.8 + value * 0.2
val = 0.4 + value * 0.6
rgb = colorsys.hsv_to_rgb(hue, sat, val)
r, g, b = [int(255 * x) for x in rgb]
if 0.3 < value < 0.6:
r = min(255, r + int(100 * value))
return (r, g, b)
def ocean(value):
if value > 0.9:
foam = int(255 * (value - 0.9) * 10)
return (foam, foam, foam)
if value < 0.5:
hue = 0.6 + value * 0.1
sat = 0.9 - value * 0.3
val = 0.3 + value * 0.7
else:
hue = 0.5 + value * 0.1
sat = 0.7
val = 0.6 + value * 0.4
rgb = colorsys.hsv_to_rgb(hue, sat, val)
return tuple(int(255 * x) for x in rgb)
def forest(value):
if value < 0.4:
hue = 0.25 + value * 0.1
sat = 0.9 - value * 0.2
val = 0.3 + value * 0.7
elif value < 0.7:
hue = 0.28 + value * 0.05
sat = 0.8
val = 0.6 + value * 0.4
else:
hue = 0.08
sat = 0.7 - (value - 0.7) * 0.5
val = 0.6 + (value - 0.7) * 0.4
rgb = colorsys.hsv_to_rgb(hue, sat, val)
return tuple(int(255 * x) for x in rgb)
def psychedelic(value):
hue = (value * 5) % 1.0
sat = 1.0
val = 0.9
rgb = colorsys.hsv_to_rgb(hue, sat, val)
r, g, b = [int(255 * x) for x in rgb]
pulse = abs(math.sin(value * math.pi * 2))
r = min(255, r + int(50 * pulse))
g = min(255, g + int(50 * pulse))
b = min(255, b + int(50 * pulse))
return (r, g, b)
color_functions = {
"Classic White-Grey": classic_white_grey,
"Electric Blue": electric_blue,
"Fire": fire,
"Rainbow": rainbow,
"Deep Space": deep_space,
"Ocean": ocean,
"Forest": forest,
"Psychedelic": psychedelic
}
return color_functions.get(preset, classic_white_grey)
def compute_fractal(self, width, height, x_min, x_max, y_min, y_max, max_iter, power, escape_radius, smooth, preset):
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
x = torch.linspace(x_min, x_max, width, device=device).view(1, width)
y = torch.linspace(y_min, y_max, height, device=device).view(height, 1)
# Create complex plane using broadcasting
real = x.expand(height, width)
imag = y.expand(height, width)
C = torch.complex(real, imag)
if preset == "Julia Set":
Z = C
C = torch.full_like(Z, -0.4 + 0.6j)
elif preset == "Burning Ship":
Z = torch.zeros_like(C)
elif preset == "Tricorn":
Z = torch.zeros_like(C)
elif preset == "Newton":
Z = C
roots = torch.tensor([1, -0.5 + 0.866j, -0.5 - 0.866j], device=device)
else: # Mandelbrot
Z = torch.zeros_like(C)
M = torch.full((height, width), max_iter, device=device, dtype=torch.float32)
for i in range(max_iter):
mask = torch.abs(Z) <= escape_radius
if preset == "Burning Ship":
Z[mask] = (torch.abs(Z[mask].real) + 1j * torch.abs(Z[mask].imag)) ** power + C[mask]
elif preset == "Tricorn":
Z[mask] = (Z[mask].conj()) ** power + C[mask]
elif preset == "Newton":
mask_newton = torch.abs(Z ** 3 - 1) > 1e-6
Z[mask_newton] = Z[mask_newton] - (Z[mask_newton] ** 3 - 1) / (3 * Z[mask_newton] ** 2)
for j, root in enumerate(roots):
close_to_root = torch.abs(Z - root) < 1e-6
M[close_to_root & (M == max_iter)] = i + j/3
else: # Mandelbrot and Julia
Z[mask] = Z[mask] ** power + C[mask]
if preset != "Newton":
escaped = mask & (torch.abs(Z) > escape_radius)
M[escaped] = i
if smooth and preset != "Newton":
abs_Z = torch.abs(Z)
outside_set = M < max_iter
smooth_M = M.float()
if outside_set.any():
valid_Z = abs_Z[outside_set]
valid_Z = torch.maximum(valid_Z, torch.tensor(1e-6, device=device))
nu = torch.zeros_like(valid_Z)
valid_mask = valid_Z > 1e-6
if valid_mask.any():
nu[valid_mask] = torch.log2(torch.log2(valid_Z[valid_mask])/math.log2(escape_radius))
smooth_M[outside_set] = M[outside_set] + 1 - nu
return (smooth_M / max_iter).cpu().numpy()
else:
return (M / max_iter).cpu().numpy()
def apply_coloring(self, fractal, width, height, color_function, color_cycles):
image = Image.new("RGB", (width, height))
pixels = image.load()
for y in range(height):
for x in range(width):
value = fractal[y, x]
if value == 1.0: # Inside set
pixels[x, y] = (0, 0, 0)
else:
cycled_value = (value * color_cycles) % 1.0
pixels[x, y] = color_function(cycled_value)
return image
def generate_fractal(self, width, height, max_iterations, preset, zoom_level, color_preset,
power, escape_radius, color_cycles, smooth_coloring, x_center=None, y_center=None):
if preset != "Custom" and x_center is None:
x_center, y_center, zoom_level = self.get_preset_coordinates(preset, zoom_level)
elif x_center is None:
x_center, y_center = -0.75, 0.0
window_size = 4.0 / zoom_level
x_min = x_center - window_size/2
x_max = x_center + window_size/2
y_min = y_center - window_size/2
y_max = y_center + window_size/2
fractal = self.compute_fractal(width, height, x_min, x_max, y_min, y_max,
max_iterations, power, escape_radius, smooth_coloring, preset)
color_function = self.get_color_function(color_preset)
image = self.apply_coloring(fractal, width, height, color_function, color_cycles)
if image.mode != 'RGB':
image = image.convert('RGB')
image_np = np.array(image).astype(np.float32) / 255.0
image_tensor = torch.from_numpy(image_np)
if len(image_tensor.shape) == 3:
image_tensor = image_tensor.unsqueeze(0)
return (image_tensor,)
# Node registration for ComfyUI
NODE_CLASS_MAPPINGS = {
"FractalGeneratorV3": FractalGeneratorV3
}
NODE_DISPLAY_NAME_MAPPINGS = {
"FractalGeneratorV3": "Fractal Gen Cuda"
}