File size: 5,079 Bytes
d195287 a547253 d195287 a547253 d195287 a547253 d195287 | 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 | from collections import OrderedDict
from typing import Dict, Iterable, List
FEATURE_VERSION = "qohlc_v2"
FEATURE_VERSION_ID = 2
WINDOW_SECONDS = 5
SEGMENT_SECONDS = 300
TOKENS_PER_SEGMENT = SEGMENT_SECONDS // WINDOW_SECONDS
LOOKBACK_SECONDS = [15, 30, 60, 120]
FEATURE_NAMES: List[str] = [
"cum_log_return",
"mean_log_return_1s",
"std_log_return_1s",
"max_up_1s",
"max_down_1s",
"realized_vol",
"window_range_frac",
"close_to_close_slope",
"accel_proxy",
"frac_pos_1s",
"frac_neg_1s",
]
for lookback in LOOKBACK_SECONDS:
prefix = f"lb_{lookback}s"
FEATURE_NAMES.extend([
f"{prefix}_dist_high",
f"{prefix}_dist_low",
f"{prefix}_drawdown_high",
f"{prefix}_rebound_low",
f"{prefix}_pos_in_range",
f"{prefix}_range_width",
f"{prefix}_compression_ratio",
f"{prefix}_breakout_high",
f"{prefix}_breakdown_low",
f"{prefix}_reclaim_breakdown",
f"{prefix}_rejection_breakout",
])
FEATURE_NAMES.extend([
"nearest_support_dist",
"nearest_resistance_dist",
"support_touch_count",
"resistance_touch_count",
"support_age_sec",
"resistance_age_sec",
"support_strength",
"resistance_strength",
"inside_support_zone",
"inside_resistance_zone",
"support_swept",
"resistance_swept",
"support_reclaim",
"resistance_reject",
"keylevel_breakout_up",
"keylevel_breakout_down",
"keylevel_hold_above",
"keylevel_hold_below",
"keylevel_failed_breakout_up",
"keylevel_failed_breakout_down",
"keylevel_flip_to_support",
"keylevel_flip_to_resistance",
"keylevel_upper_distance",
"keylevel_lower_distance",
"keylevel_zone_width_frac",
"keylevel_density",
"lower_trendline_slope",
"upper_trendline_slope",
"dist_to_lower_line",
"dist_to_upper_line",
"trend_channel_width",
"trend_convergence",
"trend_breakout_upper",
"trend_breakdown_lower",
"trend_reentry",
"ema_fast",
"ema_medium",
"sma_fast",
"sma_medium",
"price_minus_ema_fast",
"price_minus_ema_medium",
"ema_spread",
"price_zscore",
"mean_reversion_score",
"rolling_vol_zscore",
])
FEATURE_NAMES.extend([
"sr_available",
"trendline_available",
])
FEATURE_INDEX = {name: idx for idx, name in enumerate(FEATURE_NAMES)}
NUM_QUANT_OHLC_FEATURES = len(FEATURE_NAMES)
FEATURE_GROUPS = OrderedDict([
("price_path", [
"cum_log_return",
"mean_log_return_1s",
"std_log_return_1s",
"max_up_1s",
"max_down_1s",
"realized_vol",
"window_range_frac",
"close_to_close_slope",
"accel_proxy",
"frac_pos_1s",
"frac_neg_1s",
]),
("relative_structure", [name for name in FEATURE_NAMES if name.startswith("lb_")]),
("levels_breaks", [
"nearest_support_dist",
"nearest_resistance_dist",
"support_touch_count",
"resistance_touch_count",
"support_age_sec",
"resistance_age_sec",
"support_strength",
"resistance_strength",
"inside_support_zone",
"inside_resistance_zone",
"support_swept",
"resistance_swept",
"support_reclaim",
"resistance_reject",
"keylevel_breakout_up",
"keylevel_breakout_down",
"keylevel_hold_above",
"keylevel_hold_below",
"keylevel_failed_breakout_up",
"keylevel_failed_breakout_down",
"keylevel_flip_to_support",
"keylevel_flip_to_resistance",
"keylevel_upper_distance",
"keylevel_lower_distance",
"keylevel_zone_width_frac",
"keylevel_density",
]),
("trendlines", [
"lower_trendline_slope",
"upper_trendline_slope",
"dist_to_lower_line",
"dist_to_upper_line",
"trend_channel_width",
"trend_convergence",
"trend_breakout_upper",
"trend_breakdown_lower",
"trend_reentry",
]),
("rolling_quant", [
"ema_fast",
"ema_medium",
"sma_fast",
"sma_medium",
"price_minus_ema_fast",
"price_minus_ema_medium",
"ema_spread",
"price_zscore",
"mean_reversion_score",
"rolling_vol_zscore",
]),
("availability", [
"sr_available",
"trendline_available",
]),
])
def empty_feature_dict() -> Dict[str, float]:
return {name: 0.0 for name in FEATURE_NAMES}
def feature_dict_to_vector(features: Dict[str, float]) -> List[float]:
out: List[float] = []
for name in FEATURE_NAMES:
value = features.get(name, 0.0)
try:
out.append(float(value))
except Exception:
out.append(0.0)
return out
def group_feature_indices(group_names: Iterable[str]) -> List[int]:
indices: List[int] = []
for group_name in group_names:
for feature_name in FEATURE_GROUPS[group_name]:
indices.append(FEATURE_INDEX[feature_name])
return sorted(set(indices))
|