Spaces:
Runtime error
Runtime error
File size: 11,119 Bytes
d3932f4 |
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 |
"""
Dataset Generator for Candlestick Charts
Run this to create training data before training the model.
Usage:
python generate_data.py --num_samples 10000 --output_dir ./dataset
"""
import os
import json
import random
import argparse
import numpy as np
from PIL import Image
from tqdm import tqdm
import matplotlib.pyplot as plt
from matplotlib.patches import Rectangle
import io
class CandlestickGenerator:
def __init__(self, image_size=(128, 128), num_candles=20):
self.image_size = image_size
self.num_candles = num_candles
self.bg_color = "#1a1a2e"
self.bullish_color = "#00ff88"
self.bearish_color = "#ff4466"
self.patterns = {
"bullish_trend": self._bullish_trend,
"bearish_trend": self._bearish_trend,
"sideways": self._sideways,
"volatile": self._volatile,
"bullish_reversal": self._bullish_reversal,
"bearish_reversal": self._bearish_reversal,
"double_top": self._double_top,
"double_bottom": self._double_bottom,
}
def _bullish_trend(self, n, vol):
candles = []
price = 100
for i in range(n):
trend = random.uniform(0.5, 2.0)
noise = random.gauss(0, vol)
o = price + noise
c = o + random.uniform(0, vol * 2) + trend
if random.random() < 0.7:
c = max(c, o + 0.5)
h = max(o, c) + random.uniform(0, vol)
l = min(o, c) - random.uniform(0, vol)
candles.append({"o": o, "h": h, "l": l, "c": c})
price = c
return candles
def _bearish_trend(self, n, vol):
candles = []
price = 150
for i in range(n):
trend = random.uniform(0.5, 2.0)
noise = random.gauss(0, vol)
o = price + noise
c = o - random.uniform(0, vol * 2) - trend
if random.random() < 0.7:
c = min(c, o - 0.5)
h = max(o, c) + random.uniform(0, vol)
l = min(o, c) - random.uniform(0, vol)
candles.append({"o": o, "h": h, "l": l, "c": c})
price = c
return candles
def _sideways(self, n, vol):
candles = []
base = 100
for i in range(n):
center = base + random.gauss(0, vol * 2)
o = center + random.gauss(0, vol)
c = center + random.gauss(0, vol)
h = max(o, c) + random.uniform(0, vol)
l = min(o, c) - random.uniform(0, vol)
candles.append({"o": o, "h": h, "l": l, "c": c})
return candles
def _volatile(self, n, vol):
candles = []
price = 100
high_vol = vol * 3
for i in range(n):
direction = 1 if random.random() > 0.5 else -1
move = random.uniform(high_vol, high_vol * 2) * direction
o = price + random.gauss(0, high_vol)
c = o + move
h = max(o, c) + random.uniform(high_vol * 0.5, high_vol)
l = min(o, c) - random.uniform(high_vol * 0.5, high_vol)
candles.append({"o": o, "h": h, "l": l, "c": c})
price = c
return candles
def _bullish_reversal(self, n, vol):
mid = n // 2
part1 = self._bearish_trend(mid, vol)
last = part1[-1]["c"]
part2 = []
price = last
for i in range(n - mid):
trend = random.uniform(0.5, 1.5)
o = price + random.gauss(0, vol)
c = o + random.uniform(0, vol * 2) + trend
h = max(o, c) + random.uniform(0, vol)
l = min(o, c) - random.uniform(0, vol)
part2.append({"o": o, "h": h, "l": l, "c": c})
price = c
return part1 + part2
def _bearish_reversal(self, n, vol):
mid = n // 2
part1 = self._bullish_trend(mid, vol)
last = part1[-1]["c"]
part2 = []
price = last
for i in range(n - mid):
trend = random.uniform(0.5, 1.5)
o = price + random.gauss(0, vol)
c = o - random.uniform(0, vol * 2) - trend
h = max(o, c) + random.uniform(0, vol)
l = min(o, c) - random.uniform(0, vol)
part2.append({"o": o, "h": h, "l": l, "c": c})
price = c
return part1 + part2
def _double_top(self, n, vol):
third = n // 3
candles = []
base, peak = 100, 120
for i in range(third):
p = base + (peak - base) * (i / third) + random.gauss(0, vol)
o, c = p, p + random.uniform(-vol, vol)
h = max(o, c) + random.uniform(0, vol)
l = min(o, c) - random.uniform(0, vol)
candles.append({"o": o, "h": h, "l": l, "c": c})
for i in range(third):
p = peak - (peak - base) * 0.5 * (i / third) + random.gauss(0, vol)
o, c = p, p + random.uniform(-vol, vol)
h = max(o, c) + random.uniform(0, vol)
l = min(o, c) - random.uniform(0, vol)
candles.append({"o": o, "h": h, "l": l, "c": c})
for i in range(n - 2 * third):
prog = i / (n - 2 * third)
if prog < 0.5:
p = (base + peak) / 2 + (peak - (base + peak) / 2) * (prog * 2)
else:
p = peak - (peak - base) * ((prog - 0.5) * 2)
p += random.gauss(0, vol)
o, c = p, p + random.uniform(-vol, vol)
h = max(o, c) + random.uniform(0, vol)
l = min(o, c) - random.uniform(0, vol)
candles.append({"o": o, "h": h, "l": l, "c": c})
return candles
def _double_bottom(self, n, vol):
third = n // 3
candles = []
base, bottom = 120, 100
for i in range(third):
p = base - (base - bottom) * (i / third) + random.gauss(0, vol)
o, c = p, p + random.uniform(-vol, vol)
h = max(o, c) + random.uniform(0, vol)
l = min(o, c) - random.uniform(0, vol)
candles.append({"o": o, "h": h, "l": l, "c": c})
for i in range(third):
p = bottom + (base - bottom) * 0.5 * (i / third) + random.gauss(0, vol)
o, c = p, p + random.uniform(-vol, vol)
h = max(o, c) + random.uniform(0, vol)
l = min(o, c) - random.uniform(0, vol)
candles.append({"o": o, "h": h, "l": l, "c": c})
for i in range(n - 2 * third):
prog = i / (n - 2 * third)
if prog < 0.5:
p = (base + bottom) / 2 - ((base + bottom) / 2 - bottom) * (prog * 2)
else:
p = bottom + (base - bottom) * ((prog - 0.5) * 2)
p += random.gauss(0, vol)
o, c = p, p + random.uniform(-vol, vol)
h = max(o, c) + random.uniform(0, vol)
l = min(o, c) - random.uniform(0, vol)
candles.append({"o": o, "h": h, "l": l, "c": c})
return candles
def render(self, candles):
fig, ax = plt.subplots(figsize=(self.image_size[0]/100, self.image_size[1]/100), dpi=100)
fig.patch.set_facecolor(self.bg_color)
ax.set_facecolor(self.bg_color)
highs = [c["h"] for c in candles]
lows = [c["l"] for c in candles]
price_min = min(lows) * 0.98
price_max = max(highs) * 1.02
width = 0.6
for i, c in enumerate(candles):
color = self.bullish_color if c["c"] >= c["o"] else self.bearish_color
ax.plot([i, i], [c["l"], c["h"]], color=color, linewidth=1)
body_bottom = min(c["o"], c["c"])
body_height = abs(c["c"] - c["o"]) or 0.1
rect = Rectangle((i - width/2, body_bottom), width, body_height,
facecolor=color, edgecolor=color)
ax.add_patch(rect)
ax.set_xlim(-1, len(candles))
ax.set_ylim(price_min, price_max)
ax.axis("off")
buf = io.BytesIO()
plt.savefig(buf, format="png", facecolor=self.bg_color,
bbox_inches="tight", pad_inches=0.1)
plt.close(fig)
buf.seek(0)
img = Image.open(buf).convert("RGB")
img = img.resize(self.image_size, Image.Resampling.LANCZOS)
return img
def generate_sample(self):
pattern = random.choice(list(self.patterns.keys()))
vol_name = random.choice(["low", "medium", "high"])
vol_map = {"low": 1.0, "medium": 3.0, "high": 6.0}
candles = self.patterns[pattern](self.num_candles, vol_map[vol_name])
image = self.render(candles)
descriptions = {
"bullish_trend": [f"bullish trend {vol_name} volatility", f"upward trending market {vol_name} movement", "strong buying pressure"],
"bearish_trend": [f"bearish trend {vol_name} volatility", f"downward trending market {vol_name} movement", "strong selling pressure"],
"sideways": [f"sideways market {vol_name} volatility", "range-bound trading", "consolidation pattern"],
"volatile": ["highly volatile market", "erratic price movement", "choppy market conditions"],
"bullish_reversal": [f"bullish reversal {vol_name} volatility", "v-shaped recovery", "trend change bearish to bullish"],
"bearish_reversal": [f"bearish reversal {vol_name} volatility", "inverted v pattern", "trend change bullish to bearish"],
"double_top": [f"double top pattern {vol_name} volatility", "m-shaped reversal", "bearish double top"],
"double_bottom": [f"double bottom pattern {vol_name} volatility", "w-shaped reversal", "bullish double bottom"],
}
description = random.choice(descriptions[pattern])
return image, description
def generate_dataset(output_dir, num_samples=10000, image_size=128):
os.makedirs(output_dir, exist_ok=True)
os.makedirs(os.path.join(output_dir, "images"), exist_ok=True)
generator = CandlestickGenerator(image_size=(image_size, image_size))
labels = {}
print(f"Generating {num_samples} samples...")
for i in tqdm(range(num_samples)):
image, description = generator.generate_sample()
filename = f"chart_{i:06d}.png"
image.save(os.path.join(output_dir, "images", filename))
labels[filename] = description
with open(os.path.join(output_dir, "labels.json"), "w") as f:
json.dump(labels, f, indent=2)
print(f"✅ Dataset saved to {output_dir}")
print(f" - {num_samples} images")
print(f" - Labels in labels.json")
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--num_samples", type=int, default=10000)
parser.add_argument("--output_dir", type=str, default="./dataset")
parser.add_argument("--image_size", type=int, default=128)
args = parser.parse_args()
generate_dataset(args.output_dir, args.num_samples, args.image_size)
|