Datasets:
Tasks:
Tabular Classification
Formats:
parquet
Languages:
English
Size:
< 1K
Tags:
economics
quantitative-finance
causal-inference
macroeconomics
housing-economics
market-microstructure
License:
File size: 8,453 Bytes
ebcde1f | 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 | from __future__ import annotations
import polars as pl
import pytest
from microstructure.research.analysis import (
LiquidityShockThresholds,
RegimeThresholds,
assign_market_regimes,
cross_instrument_stability_summary,
estimate_signal_half_life,
feature_stability_summary,
intraday_liquidity_summary,
large_trade_price_impact_summary,
liquidity_recovery_summary,
ofi_future_return_association,
regime_outcome_summary,
)
MINUTE = 60_000_000_000
def test_intraday_liquidity_uses_fixed_reproducible_buckets() -> None:
frame = pl.DataFrame(
{
"symbol": ["BTCUSDT"] * 4,
"decision_ts_ns": [0, 30 * MINUTE, 60 * MINUTE, 90 * MINUTE],
"spread_bps": [2.0, 4.0, 6.0, 8.0],
"depth_total_l1": [100.0, 80.0, 60.0, 40.0],
"queue_imbalance_l1": [0.2, 0.0, -0.2, 0.4],
}
)
summary = intraday_liquidity_summary(frame, bucket_minutes=60)
first = summary.row(0, named=True)
assert first["intraday_bucket_label"] == "00:00"
assert first["n_observations"] == 2
assert first["mean_spread_bps"] == 3.0
assert first["mean_depth_l1"] == 90.0
assert summary.get_column("descriptive_only").all()
def test_ofi_association_and_half_life_use_supplied_horizons() -> None:
ofi = [-3.0, -2.0, -1.0, 0.0, 1.0, 2.0, 3.0, 4.0]
orthogonal_noise = [1.0, -1.0, -1.0, 1.0, 1.0, -1.0, -1.0, 1.0]
frame = pl.DataFrame(
{
"symbol": ["BTCUSDT"] * len(ofi),
"ofi_l1": ofi,
"return_h1": ofi,
"return_h2": [
value + noise for value, noise in zip(ofi, orthogonal_noise, strict=True)
],
"return_h4": orthogonal_noise,
}
)
association = ofi_future_return_association(
frame,
horizon_return_columns={1: "return_h1", 2: "return_h2", 4: "return_h4"},
)
at_one = association.filter(pl.col("horizon_events") == 1).row(0, named=True)
assert at_one["pearson_correlation"] == pytest.approx(1.0)
assert at_one["ols_slope_return_per_ofi_unit"] == pytest.approx(1.0)
assert at_one["descriptive_only"] is True
half_life = estimate_signal_half_life(association)
result = half_life.summary.row(0, named=True)
assert result["reference_horizon_events"] == 1
assert result["first_crossing_half_life_events"] == 4.0
assert result["analysis_kind"] == "signal_half_life_descriptive"
assert half_life.curve.get_column("normalized_absolute_correlation")[0] == pytest.approx(1.0)
def test_large_trade_impact_requires_caller_supplied_train_threshold() -> None:
frame = pl.DataFrame(
{
"symbol": ["BTCUSDT"] * 4,
"quantity": [1.0, 2.0, 10.0, 20.0],
"impact_h2": [1.0, 2.0, 10.0, 20.0],
}
)
summary = large_trade_price_impact_summary(
frame,
impact_columns={2: "impact_h2"},
train_quantity_thresholds={"BTCUSDT": 5.0},
)
regular = summary.filter(~pl.col("large_trade")).row(0, named=True)
large = summary.filter(pl.col("large_trade")).row(0, named=True)
assert regular["mean_signed_impact_bps"] == 1.5
assert large["mean_signed_impact_bps"] == 15.0
assert large["train_quantity_threshold"] == 5.0
assert large["threshold_source"] == "caller_supplied_train_period"
def test_liquidity_recovery_tracks_one_episode_and_censors_segment_tail() -> None:
frame = pl.DataFrame(
{
"symbol": ["BTCUSDT"] * 6,
"continuity_id": ["a"] * 6,
"decision_ts_ns": list(range(6)),
"decision_sequence": list(range(1, 7)),
"spread_bps": [2.0, 10.0, 8.0, 4.0, 3.0, 9.0],
"depth_total_l1": [100.0, 40.0, 60.0, 90.0, 100.0, 30.0],
}
)
summary = liquidity_recovery_summary(
frame,
train_thresholds={
"BTCUSDT": LiquidityShockThresholds(
spread_shock_bps=8.0,
depth_shock_max=50.0,
spread_recovery_bps=4.0,
depth_recovery_min=80.0,
max_recovery_events=3,
)
},
)
assert summary.height == 2
recovered = summary.row(0, named=True)
assert recovered["shock_sequence"] == 2
assert recovered["recovery_events"] == 2
assert recovered["recovery_time_ns"] == 2
assert recovered["recovery_right_censored"] is False
assert recovered["threshold_source"] == "caller_supplied_train_period"
tail = summary.row(1, named=True)
assert tail["shock_sequence"] == 6
assert tail["recovered"] is None
assert tail["recovery_right_censored"] is True
assert tail["recovery_information_end_ts_ns"] is None
def test_liquidity_recovery_infers_late_censor_status_without_row_limit() -> None:
row_count = 201
frame = pl.DataFrame(
{
"symbol": ["BTCUSDT"] * row_count,
"continuity_id": ["a"] * row_count,
"decision_ts_ns": list(range(row_count)),
"decision_sequence": list(range(row_count)),
"spread_bps": [10.0 if index % 2 == 0 else 2.0 for index in range(row_count)],
"depth_total_l1": [40.0 if index % 2 == 0 else 100.0 for index in range(row_count)],
}
)
summary = liquidity_recovery_summary(
frame,
train_thresholds={
"BTCUSDT": LiquidityShockThresholds(
spread_shock_bps=8.0,
depth_shock_max=50.0,
spread_recovery_bps=4.0,
depth_recovery_min=80.0,
max_recovery_events=1,
)
},
)
assert summary.height == 101
assert summary.get_column("recovery_censor_reason")[-1] == ("segment_ends_before_max_horizon")
def test_regimes_are_assigned_from_supplied_train_boundaries() -> None:
frame = pl.DataFrame(
{
"symbol": ["BTCUSDT"] * 3,
"volatility": [0.1, 0.5, 0.9],
"spread_bps": [1.0, 3.0, 6.0],
"depth_total_l1": [120.0, 80.0, 40.0],
"future_return": [0.01, 0.0, -0.02],
}
)
thresholds = {
"BTCUSDT": RegimeThresholds(
volatility_low=0.2,
volatility_high=0.8,
spread_tight_bps=2.0,
spread_wide_bps=5.0,
depth_low=50.0,
depth_high=100.0,
)
}
regimes = assign_market_regimes(
frame, train_thresholds=thresholds, volatility_column="volatility"
)
assert regimes.get_column("joint_market_regime").to_list() == [
"low__liquid",
"medium__normal",
"high__stressed",
]
assert regimes.get_column("regime_threshold_source").unique().to_list() == [
"caller_supplied_train_period"
]
outcomes = regime_outcome_summary(regimes, outcome_columns=("future_return",))
assert outcomes.get_column("n_observations").sum() == 3
assert outcomes.get_column("descriptive_only").all()
def test_cross_instrument_and_feature_stability_are_descriptive() -> None:
effects = pl.DataFrame(
{
"symbol": ["BTCUSDT", "ETHUSDT", "BTCUSDT", "ETHUSDT"],
"horizon_events": [1, 1, 2, 2],
"effect": [0.2, 0.1, -0.1, -0.2],
}
)
cross = cross_instrument_stability_summary(effects, value_column="effect")
assert cross.get_column("sign_agreement_fraction").to_list() == [1.0, 1.0]
assert cross.get_column("n_instruments").to_list() == [2, 2]
reference = pl.DataFrame(
{
"symbol": ["BTCUSDT"] * 4,
"stable": [0.0, 1.0, 2.0, 3.0],
"shifted": [0.0, 1.0, 2.0, 3.0],
}
)
comparison = pl.DataFrame(
{
"symbol": ["BTCUSDT"] * 4,
"stable": [0.0, 1.0, 2.0, 3.0],
"shifted": [2.0, 3.0, 4.0, 5.0],
}
)
stability = feature_stability_summary(
reference,
comparison,
feature_columns=("stable", "shifted"),
bins=2,
)
stable = stability.filter(pl.col("feature") == "stable").row(0, named=True)
shifted = stability.filter(pl.col("feature") == "shifted").row(0, named=True)
assert stable["population_stability_index"] == pytest.approx(0.0)
assert shifted["population_stability_index"] > 0
assert shifted["standardized_mean_shift"] > 1.0
assert shifted["bin_source"] == "reference_period_only"
assert shifted["descriptive_only"] is True
|