wehub-resource-sync
4715 lines
438 KiB
Plaintext
4715 lines
438 KiB
Plaintext
{
|
||
"cells": [
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "f57dc948",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"# Feature Evaluation — ETFs\n",
|
||
"\n",
|
||
"Consolidated evaluation of Ch8 financial features and Ch9 temporal features\n",
|
||
"against forward return labels. Produces triage decisions for Ch11 modeling.\n",
|
||
"\n",
|
||
"**Learning Objectives**:\n",
|
||
"- Evaluate individual feature predictive power via Information Coefficient (IC)\n",
|
||
"- Apply HAC adjustment for overlapping-return autocorrelation\n",
|
||
"- Control false discovery rate with Benjamini-Hochberg correction\n",
|
||
"- Assess feature redundancy and family-level signal concentration\n",
|
||
"- Produce a triage ledger for downstream model selection\n",
|
||
"\n",
|
||
"**Book Reference**: Chapter 8, Section 8.5 (Feature Evaluation)\n",
|
||
"\n",
|
||
"**Prerequisites**: `03_financial_features.py` and `04_temporal.py` must have run\n",
|
||
"(produces `features/financial.parquet` and `features/model_based.parquet`)."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 1,
|
||
"id": "f466e1e6",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:21:16.233733Z",
|
||
"iopub.status.busy": "2026-05-15T05:21:16.233295Z",
|
||
"iopub.status.idle": "2026-05-15T05:21:24.013776Z",
|
||
"shell.execute_reply": "2026-05-15T05:21:24.012479Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"\"\"\"Feature Evaluation - ETFs case study.\"\"\"\n",
|
||
"\n",
|
||
"import warnings\n",
|
||
"from pathlib import Path\n",
|
||
"\n",
|
||
"warnings.filterwarnings(\"ignore\")\n",
|
||
"\n",
|
||
"import numpy as np\n",
|
||
"import plotly.graph_objects as go\n",
|
||
"import polars as pl\n",
|
||
"import yaml\n",
|
||
"from ml4t.diagnostic.evaluation.stats import benjamini_hochberg_fdr\n",
|
||
"from ml4t.diagnostic.metrics import compute_ic_hac_stats, cross_sectional_ic_series\n",
|
||
"from plotly.subplots import make_subplots\n",
|
||
"from scipy.stats import spearmanr\n",
|
||
"\n",
|
||
"from utils.paths import get_case_study_dir"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 2,
|
||
"id": "e19eb193",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:21:24.030676Z",
|
||
"iopub.status.busy": "2026-05-15T05:21:24.030223Z",
|
||
"iopub.status.idle": "2026-05-15T05:21:24.035290Z",
|
||
"shell.execute_reply": "2026-05-15T05:21:24.034521Z"
|
||
},
|
||
"tags": [
|
||
"parameters"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"# Production defaults\n",
|
||
"MAX_SYMBOLS = 0"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 3,
|
||
"id": "53a1a626",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:21:24.045481Z",
|
||
"iopub.status.busy": "2026-05-15T05:21:24.045133Z",
|
||
"iopub.status.idle": "2026-05-15T05:21:24.051183Z",
|
||
"shell.execute_reply": "2026-05-15T05:21:24.050036Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"CASE_STUDY_ID = \"etfs\"\n",
|
||
"CASE_DIR = get_case_study_dir(CASE_STUDY_ID)\n",
|
||
"EVAL_DIR = CASE_DIR / \"evaluation\"\n",
|
||
"EVAL_DIR.mkdir(exist_ok=True)\n",
|
||
"\n",
|
||
"# ETFs config\n",
|
||
"PRIMARY_LABEL_FILE = \"fwd_ret_21d.parquet\"\n",
|
||
"HAC_MAXLAGS = 21 # 21-day forward return\n",
|
||
"MIN_PERIODS = 10 # 99 symbols\n",
|
||
"IC_THRESHOLD = 0.01 # Monthly horizon"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "9e636e00",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"## 0. Load Artifacts & Build Evaluation Panel\n",
|
||
"\n",
|
||
"We load the pre-computed feature matrices and labels, then join into a single\n",
|
||
"evaluation panel. Temporal features include both date-level (HMM regime, FFD)\n",
|
||
"and per-symbol (GARCH) features, so we join on `[date, symbol]`."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 4,
|
||
"id": "030da08e",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:21:24.071082Z",
|
||
"iopub.status.busy": "2026-05-15T05:21:24.070648Z",
|
||
"iopub.status.idle": "2026-05-15T05:21:24.955584Z",
|
||
"shell.execute_reply": "2026-05-15T05:21:24.951423Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Features: (396186, 59)\n",
|
||
"Temporal: (470662, 16)\n",
|
||
"Labels: (468562, 3), column: fwd_ret_21d\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Load features (normalize date column type to pl.Date for consistent joins)\n",
|
||
"features = pl.read_parquet(CASE_DIR / \"features\" / \"financial.parquet\").with_columns(\n",
|
||
" pl.col(\"timestamp\").cast(pl.Date)\n",
|
||
")\n",
|
||
"temporal = pl.read_parquet(CASE_DIR / \"features\" / \"model_based.parquet\").with_columns(\n",
|
||
" pl.col(\"timestamp\").cast(pl.Date)\n",
|
||
")\n",
|
||
"# model_based.parquet now carries one row per (timestamp, symbol, fold);\n",
|
||
"# drop fold and de-dup so the join doesn't multiply the eval panel by fold count.\n",
|
||
"if \"fold\" in temporal.columns:\n",
|
||
" temporal = temporal.drop(\"fold\").unique(subset=[\"timestamp\", \"symbol\"], keep=\"last\")\n",
|
||
"\n",
|
||
"# Load primary label\n",
|
||
"label_df = pl.read_parquet(CASE_DIR / \"labels\" / PRIMARY_LABEL_FILE).with_columns(\n",
|
||
" pl.col(\"timestamp\").cast(pl.Date)\n",
|
||
")\n",
|
||
"label_col = [c for c in label_df.columns if c not in (\"timestamp\", \"symbol\")][0]\n",
|
||
"\n",
|
||
"# Load evaluation config from setup.yaml\n",
|
||
"with open(CASE_DIR / \"config\" / \"setup.yaml\") as f:\n",
|
||
" setup_config = yaml.safe_load(f)\n",
|
||
"eval_config = setup_config.get(\"evaluation\", {})\n",
|
||
"\n",
|
||
"print(f\"Features: {features.shape}\")\n",
|
||
"print(f\"Temporal: {temporal.shape}\")\n",
|
||
"print(f\"Labels: {label_df.shape}, column: {label_col}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 5,
|
||
"id": "671ad81a",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:21:24.969977Z",
|
||
"iopub.status.busy": "2026-05-15T05:21:24.969515Z",
|
||
"iopub.status.idle": "2026-05-15T05:21:25.278995Z",
|
||
"shell.execute_reply": "2026-05-15T05:21:25.277717Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"\n",
|
||
"Eval panel: 394,233 rows, 99 symbols, 4,759 dates\n",
|
||
"Features: 57 financial + 14 temporal = 71 total\n",
|
||
"Label: fwd_ret_21d\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Identify feature columns by source\n",
|
||
"JOIN_COLS = [\"timestamp\", \"symbol\"]\n",
|
||
"DATE_COL = \"timestamp\"\n",
|
||
"\n",
|
||
"financial_cols = [c for c in features.columns if c not in JOIN_COLS]\n",
|
||
"temporal_cols = [c for c in temporal.columns if c not in (\"timestamp\", \"symbol\")]\n",
|
||
"\n",
|
||
"# Join: features + temporal (on [date, symbol]) + labels\n",
|
||
"eval_panel = features.join(temporal, on=JOIN_COLS, how=\"left\")\n",
|
||
"eval_panel = eval_panel.join(label_df, on=JOIN_COLS, how=\"inner\")\n",
|
||
"\n",
|
||
"all_feature_cols = financial_cols + temporal_cols\n",
|
||
"\n",
|
||
"if MAX_SYMBOLS > 0:\n",
|
||
" top = eval_panel.group_by(\"symbol\").len().sort(\"len\", descending=True).head(MAX_SYMBOLS)\n",
|
||
" eval_panel = eval_panel.filter(pl.col(\"symbol\").is_in(top[\"symbol\"]))\n",
|
||
"\n",
|
||
"n_rows = len(eval_panel)\n",
|
||
"n_symbols = eval_panel[\"symbol\"].n_unique()\n",
|
||
"n_dates = eval_panel[DATE_COL].n_unique()\n",
|
||
"print(f\"\\nEval panel: {n_rows:,} rows, {n_symbols} symbols, {n_dates:,} dates\")\n",
|
||
"print(\n",
|
||
" f\"Features: {len(financial_cols)} financial + {len(temporal_cols)} temporal\"\n",
|
||
" f\" = {len(all_feature_cols)} total\"\n",
|
||
")\n",
|
||
"print(f\"Label: {label_col}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "4b6f6b5e",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"## 0.5 Data Quality Gate\n",
|
||
"\n",
|
||
"Verify upstream artifacts are free of critical defects (negative prices,\n",
|
||
"infinities, extreme returns) before any statistical evaluation."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 6,
|
||
"id": "ec9ea315",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:21:25.297234Z",
|
||
"iopub.status.busy": "2026-05-15T05:21:25.296819Z",
|
||
"iopub.status.idle": "2026-05-15T05:21:25.603325Z",
|
||
"shell.execute_reply": "2026-05-15T05:21:25.598637Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Data Quality Gate: 0 CRITICAL, 1 WARNING\n",
|
||
" [!] WARNING: 2 features have values |x| > 1e+06: sharpe_5d(172), sharpe_10d(37)\n"
|
||
]
|
||
},
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"{'issues': ['WARNING: 2 features have values |x| > 1e+06: sharpe_5d(172), sharpe_10d(37)'],\n",
|
||
" 'n_critical': 0,\n",
|
||
" 'n_warning': 1}"
|
||
]
|
||
},
|
||
"execution_count": 6,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"from utils.data_quality import validate_modeling_inputs\n",
|
||
"\n",
|
||
"validate_modeling_inputs(\n",
|
||
" features_df=eval_panel,\n",
|
||
" label_df=eval_panel,\n",
|
||
" feature_cols=all_feature_cols,\n",
|
||
" label_col=label_col,\n",
|
||
" join_cols=JOIN_COLS,\n",
|
||
" asset_col=\"symbol\",\n",
|
||
" max_abs_return=1.0, # 21-day ETF returns (max observed ~0.71)\n",
|
||
" fail_on_critical=True,\n",
|
||
")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "968684f8",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"## 1. Correctness Screens\n",
|
||
"\n",
|
||
"Before evaluating predictive power, we check data quality:\n",
|
||
"- **Coverage**: fraction of non-null values (threshold: 70%)\n",
|
||
"- **Staleness**: fraction of unchanged values from prior date (threshold: 50%)\n",
|
||
"\n",
|
||
"Features that fail either gate are marked STOP in the triage."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 7,
|
||
"id": "e6e20a9c",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:21:25.618707Z",
|
||
"iopub.status.busy": "2026-05-15T05:21:25.618366Z",
|
||
"iopub.status.idle": "2026-05-15T05:21:32.863331Z",
|
||
"shell.execute_reply": "2026-05-15T05:21:32.860590Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Correctness gate: 67 PASS, 4 FAIL\n",
|
||
"shape: (4, 3)\n",
|
||
"┌──────────────────┬──────────┬───────────┐\n",
|
||
"│ feature ┆ coverage ┆ staleness │\n",
|
||
"│ --- ┆ --- ┆ --- │\n",
|
||
"│ str ┆ f64 ┆ f64 │\n",
|
||
"╞══════════════════╪══════════╪═══════════╡\n",
|
||
"│ aroon_diff ┆ 1.0 ┆ 0.589 │\n",
|
||
"│ pct_positive_63d ┆ 1.0 ┆ 0.51 │\n",
|
||
"│ vol_63d_rank ┆ 1.0 ┆ 0.596 │\n",
|
||
"│ regime ┆ 1.0 ┆ 0.983 │\n",
|
||
"└──────────────────┴──────────┴───────────┘\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"coverage = {}\n",
|
||
"staleness = {}\n",
|
||
"\n",
|
||
"for feat in all_feature_cols:\n",
|
||
" col = eval_panel[feat]\n",
|
||
" coverage[feat] = col.drop_nulls().len() / n_rows\n",
|
||
"\n",
|
||
" # Staleness: fraction unchanged from prior row within same symbol\n",
|
||
" unchanged = (\n",
|
||
" eval_panel.sort(JOIN_COLS)\n",
|
||
" .select((pl.col(feat) == pl.col(feat).shift(1).over(\"symbol\")).alias(\"same\"))[\"same\"]\n",
|
||
" .sum()\n",
|
||
" )\n",
|
||
" staleness[feat] = float(unchanged) / max(n_rows - n_symbols, 1)\n",
|
||
"\n",
|
||
"correctness = {\n",
|
||
" feat: coverage[feat] >= 0.70 and staleness[feat] <= 0.50 for feat in all_feature_cols\n",
|
||
"}\n",
|
||
"n_pass = sum(correctness.values())\n",
|
||
"n_fail = len(correctness) - n_pass\n",
|
||
"print(f\"Correctness gate: {n_pass} PASS, {n_fail} FAIL\")\n",
|
||
"\n",
|
||
"if n_fail > 0:\n",
|
||
" fail_df = pl.DataFrame(\n",
|
||
" {\n",
|
||
" \"feature\": [f for f, ok in correctness.items() if not ok],\n",
|
||
" \"coverage\": [round(coverage[f], 3) for f, ok in correctness.items() if not ok],\n",
|
||
" \"staleness\": [round(staleness[f], 3) for f, ok in correctness.items() if not ok],\n",
|
||
" }\n",
|
||
" )\n",
|
||
" print(fail_df)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "c8119a8c",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"### Effective Sample Size\n",
|
||
"\n",
|
||
"With daily observations and 21-day forward return labels, consecutive labels\n",
|
||
"overlap by 20 days. This autocorrelation reduces the effective number of\n",
|
||
"independent observations:\n",
|
||
"\n",
|
||
"$$N_{\\text{eff}} \\approx \\frac{N_{\\text{dates}}}{h} \\times N_{\\text{symbols}}$$\n",
|
||
"\n",
|
||
"where $h = 21$ is the label horizon. HAC standard errors (below) account for\n",
|
||
"this, but raw row counts overstate statistical power."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 8,
|
||
"id": "b1b03edb",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:21:32.886764Z",
|
||
"iopub.status.busy": "2026-05-15T05:21:32.886373Z",
|
||
"iopub.status.idle": "2026-05-15T05:21:32.894415Z",
|
||
"shell.execute_reply": "2026-05-15T05:21:32.892893Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Raw panel: 394,233 rows (4,759 dates × 99 symbols)\n",
|
||
"Effective sample size: ~22,374 (226 independent date blocks × 99 symbols)\n",
|
||
"Overlap reduction factor: 21x\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"n_eff_dates = n_dates // HAC_MAXLAGS\n",
|
||
"n_eff = n_eff_dates * n_symbols\n",
|
||
"print(f\"Raw panel: {n_rows:,} rows ({n_dates:,} dates × {n_symbols} symbols)\")\n",
|
||
"print(\n",
|
||
" f\"Effective sample size: ~{n_eff:,} ({n_eff_dates:,} independent date blocks × {n_symbols} symbols)\"\n",
|
||
")\n",
|
||
"print(f\"Overlap reduction factor: {n_dates / n_eff_dates:.0f}x\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "5d41079d",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"## 2. Univariate Association (IC + HAC)\n",
|
||
"\n",
|
||
"For each feature that passes correctness, we compute:\n",
|
||
"- **IC time series**: daily cross-sectional Spearman rank correlation with the label\n",
|
||
"- **HAC-adjusted t-statistics**: Newey-West standard errors with bandwidth = 21\n",
|
||
" (matching the 21-day label horizon) to account for overlapping-return autocorrelation\n",
|
||
"\n",
|
||
"Date-level features (identical across all symbols on a given date) produce zero\n",
|
||
"cross-sectional IC by construction. We detect and flag these separately."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 9,
|
||
"id": "d064773d",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:21:32.913235Z",
|
||
"iopub.status.busy": "2026-05-15T05:21:32.912941Z",
|
||
"iopub.status.idle": "2026-05-15T05:21:32.978894Z",
|
||
"shell.execute_reply": "2026-05-15T05:21:32.975857Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Date-level features (zero CS variance): ['corr_spy_tlt_63d', 'ffd_eem', 'ffd_efa', 'ffd_gld', 'ffd_hyg', 'ffd_iwm', 'ffd_lqd', 'ffd_qqq', 'ffd_spy', 'ffd_tlt', 'ffd_vnq', 'regime_log_duration', 'regime_prob_stress', 'regime_transition', 'yield_curve_slope', 'yield_curve_zscore']\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"evaluable_features = [f for f in all_feature_cols if correctness[f]]\n",
|
||
"\n",
|
||
"# Detect date-level features (zero cross-sectional variance)\n",
|
||
"# Compute std per date for all features at once, then check which have ~zero mean std\n",
|
||
"cs_std_df = eval_panel.group_by(DATE_COL).agg(\n",
|
||
" [pl.col(f).std().alias(f) for f in evaluable_features]\n",
|
||
")\n",
|
||
"date_level_features = set()\n",
|
||
"for feat in evaluable_features:\n",
|
||
" mean_std = cs_std_df[feat].drop_nulls().mean()\n",
|
||
" if mean_std is not None and mean_std < 1e-10:\n",
|
||
" date_level_features.add(feat)\n",
|
||
"\n",
|
||
"if date_level_features:\n",
|
||
" print(f\"Date-level features (zero CS variance): {sorted(date_level_features)}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 10,
|
||
"id": "7ed1b887",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:21:32.989185Z",
|
||
"iopub.status.busy": "2026-05-15T05:21:32.989020Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:01.474782Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:01.474308Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
" IC progress: 1000/4759 dates\n"
|
||
]
|
||
},
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
" IC progress: 2000/4759 dates\n"
|
||
]
|
||
},
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
" IC progress: 3000/4759 dates\n"
|
||
]
|
||
},
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
" IC progress: 4000/4759 dates\n"
|
||
]
|
||
},
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
" IC progress: 4759/4759 dates (done)\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Batch IC computation: one pass over dates, all features at once\n",
|
||
"# This is much faster than calling cross_sectional_ic_series per feature\n",
|
||
"cs_features = [f for f in evaluable_features if f not in date_level_features]\n",
|
||
"cols_needed = [DATE_COL] + cs_features + [label_col]\n",
|
||
"eval_sub = eval_panel.select(cols_needed).drop_nulls(subset=[label_col])\n",
|
||
"\n",
|
||
"# Group by date and compute Spearman IC for all features per date\n",
|
||
"dates_list = eval_sub[DATE_COL].unique().sort().to_list()\n",
|
||
"n_total = len(dates_list)\n",
|
||
"\n",
|
||
"# Pre-allocate: dict of feature -> list of (date, ic, n_obs)\n",
|
||
"ic_series_data = {feat: [] for feat in cs_features}\n",
|
||
"\n",
|
||
"for i, dt in enumerate(dates_list):\n",
|
||
" cross_section = eval_sub.filter(pl.col(DATE_COL) == dt)\n",
|
||
" n_obs = len(cross_section)\n",
|
||
" if n_obs < MIN_PERIODS:\n",
|
||
" continue\n",
|
||
"\n",
|
||
" label_arr = cross_section[label_col].to_numpy()\n",
|
||
" label_valid = ~np.isnan(label_arr)\n",
|
||
"\n",
|
||
" for feat in cs_features:\n",
|
||
" feat_arr = cross_section[feat].to_numpy()\n",
|
||
" valid_mask = label_valid & ~np.isnan(feat_arr)\n",
|
||
" n_valid = int(valid_mask.sum())\n",
|
||
" if n_valid >= MIN_PERIODS:\n",
|
||
" ic_val, _ = spearmanr(feat_arr[valid_mask], label_arr[valid_mask])\n",
|
||
" if not np.isnan(ic_val):\n",
|
||
" ic_series_data[feat].append((dt, float(ic_val), n_valid))\n",
|
||
"\n",
|
||
" if (i + 1) % 1000 == 0:\n",
|
||
" print(f\" IC progress: {i + 1}/{n_total} dates\")\n",
|
||
"\n",
|
||
"print(f\" IC progress: {n_total}/{n_total} dates (done)\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 11,
|
||
"id": "037047c3",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:01.486643Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:01.486372Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:02.087254Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:02.085008Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"IC computed for 51 cross-sectional features\n",
|
||
"Skipped 16 date-level features\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Convert to DataFrames and compute HAC stats\n",
|
||
"ic_results = {}\n",
|
||
"ic_timeseries = {}\n",
|
||
"for feat in cs_features:\n",
|
||
" data = ic_series_data[feat]\n",
|
||
" if len(data) < 20:\n",
|
||
" continue\n",
|
||
" dates_f, ics_f, nobs_f = zip(*data, strict=False)\n",
|
||
" ic_df = pl.DataFrame({DATE_COL: list(dates_f), \"ic\": list(ics_f), \"n_obs\": list(nobs_f)})\n",
|
||
" hac_stats = compute_ic_hac_stats(ic_df, ic_col=\"ic\", maxlags=HAC_MAXLAGS)\n",
|
||
" ic_results[feat] = hac_stats\n",
|
||
" ic_timeseries[feat] = ic_df\n",
|
||
"\n",
|
||
"print(f\"IC computed for {len(ic_results)} cross-sectional features\")\n",
|
||
"print(f\"Skipped {len(date_level_features)} date-level features\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "d3d9a668",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"### Fold-Level Stability\n",
|
||
"\n",
|
||
"We divide the evaluation period into approximate annual folds (matching the\n",
|
||
"CV config's 1-year test windows) and check whether IC sign is consistent\n",
|
||
"across folds. Features with > 60% positive-IC folds are more robust."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 12,
|
||
"id": "f4c4ee7d",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:02.108716Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:02.108128Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:02.205537Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:02.204487Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Fold stability computed for 51 features\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Generate approximate fold boundaries\n",
|
||
"all_dates = eval_panel[DATE_COL].unique().sort().to_list()\n",
|
||
"n_folds = eval_config.get(\"n_splits\", 8)\n",
|
||
"fold_size = len(all_dates) // n_folds\n",
|
||
"\n",
|
||
"fold_boundaries = []\n",
|
||
"for i in range(n_folds):\n",
|
||
" start_idx = i * fold_size\n",
|
||
" end_idx = min((i + 1) * fold_size - 1, len(all_dates) - 1)\n",
|
||
" fold_boundaries.append((all_dates[start_idx], all_dates[end_idx]))\n",
|
||
"\n",
|
||
"fold_stats = {}\n",
|
||
"for feat in ic_results:\n",
|
||
" fold_ics = []\n",
|
||
" ts = ic_timeseries[feat]\n",
|
||
" for fold_start, fold_end in fold_boundaries:\n",
|
||
" fold_ic = ts.filter((pl.col(DATE_COL) >= fold_start) & (pl.col(DATE_COL) <= fold_end))\n",
|
||
" if len(fold_ic) >= 5:\n",
|
||
" fold_ics.append(float(fold_ic[\"ic\"].mean()))\n",
|
||
"\n",
|
||
" if fold_ics:\n",
|
||
" sign_consistency = sum(1 for ic in fold_ics if ic > 0) / len(fold_ics)\n",
|
||
" fold_stats[feat] = {\n",
|
||
" \"n_folds\": len(fold_ics),\n",
|
||
" \"sign_consistency\": sign_consistency,\n",
|
||
" \"worst_fold_ic\": min(fold_ics),\n",
|
||
" \"best_fold_ic\": max(fold_ics),\n",
|
||
" \"median_fold_ic\": float(np.median(fold_ics)),\n",
|
||
" }\n",
|
||
"\n",
|
||
"print(f\"Fold stability computed for {len(fold_stats)} features\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "509f412d",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"## 3. Multiple Testing (BH-FDR)\n",
|
||
"\n",
|
||
"Testing 50+ features simultaneously at $\\alpha = 0.05$ expects ~2.5 false\n",
|
||
"positives. We apply the Benjamini-Hochberg procedure to control the false\n",
|
||
"discovery rate. The **inflation factor** measures how much naive significance\n",
|
||
"overstates true significance:\n",
|
||
"\n",
|
||
"$$\\text{Inflation} = \\frac{N_{\\text{naive significant}}}{N_{\\text{FDR significant}}}$$"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 13,
|
||
"id": "37a10460",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:02.238339Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:02.237784Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:02.251795Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:02.250671Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Features tested: 51\n",
|
||
"Naive significant (p < 0.05): 13\n",
|
||
"HAC significant (|t| > 1.96): 13\n",
|
||
"FDR significant (q < 0.05): 1\n",
|
||
"Inflation factor (HAC): 1.00x\n",
|
||
"Inflation factor (FDR): 13.00x\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"feature_names = list(ic_results.keys())\n",
|
||
"p_values = [ic_results[f][\"p_value\"] for f in feature_names]\n",
|
||
"\n",
|
||
"fdr_result = benjamini_hochberg_fdr(p_values, alpha=0.05, return_details=True)\n",
|
||
"\n",
|
||
"eval_summary = pl.DataFrame(\n",
|
||
" {\n",
|
||
" \"feature\": feature_names,\n",
|
||
" \"source\": [\"temporal\" if f in temporal_cols else \"financial\" for f in feature_names],\n",
|
||
" \"ic_mean\": [ic_results[f][\"mean_ic\"] for f in feature_names],\n",
|
||
" \"hac_se\": [ic_results[f][\"hac_se\"] for f in feature_names],\n",
|
||
" \"hac_t\": [ic_results[f][\"t_stat\"] for f in feature_names],\n",
|
||
" \"hac_p\": p_values,\n",
|
||
" \"fdr_p\": [float(p) for p in fdr_result[\"adjusted_p_values\"]],\n",
|
||
" \"fdr_sig\": [bool(r) for r in fdr_result[\"rejected\"]],\n",
|
||
" \"naive_t\": [ic_results[f][\"naive_t_stat\"] for f in feature_names],\n",
|
||
" }\n",
|
||
").sort(pl.col(\"ic_mean\").cast(pl.Float64, strict=False).abs(), descending=True)\n",
|
||
"\n",
|
||
"n_significant_naive = sum(1 for p in p_values if p < 0.05)\n",
|
||
"n_significant_hac = sum(1 for f in feature_names if abs(ic_results[f][\"t_stat\"]) > 1.96)\n",
|
||
"n_significant_fdr = int(fdr_result[\"n_rejected\"])\n",
|
||
"\n",
|
||
"inflation_hac = n_significant_naive / max(n_significant_hac, 1)\n",
|
||
"inflation_fdr = n_significant_naive / max(n_significant_fdr, 1)\n",
|
||
"\n",
|
||
"print(f\"Features tested: {len(feature_names)}\")\n",
|
||
"print(f\"Naive significant (p < 0.05): {n_significant_naive}\")\n",
|
||
"print(f\"HAC significant (|t| > 1.96): {n_significant_hac}\")\n",
|
||
"print(f\"FDR significant (q < 0.05): {n_significant_fdr}\")\n",
|
||
"print(f\"Inflation factor (HAC): {inflation_hac:.2f}x\")\n",
|
||
"print(f\"Inflation factor (FDR): {inflation_fdr:.2f}x\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 14,
|
||
"id": "ce9197b3",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:02.265394Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:02.264867Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:02.272596Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:02.270946Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"\n",
|
||
"Top 15 features by absolute IC:\n",
|
||
"shape: (15, 9)\n",
|
||
"┌───────────────┬───────────┬───────────┬──────────┬───┬──────────┬──────────┬─────────┬───────────┐\n",
|
||
"│ feature ┆ source ┆ ic_mean ┆ hac_se ┆ … ┆ hac_p ┆ fdr_p ┆ fdr_sig ┆ naive_t │\n",
|
||
"│ --- ┆ --- ┆ --- ┆ --- ┆ ┆ --- ┆ --- ┆ --- ┆ --- │\n",
|
||
"│ str ┆ str ┆ f64 ┆ f64 ┆ ┆ f64 ┆ f64 ┆ bool ┆ f64 │\n",
|
||
"╞═══════════════╪═══════════╪═══════════╪══════════╪═══╪══════════╪══════════╪═════════╪═══════════╡\n",
|
||
"│ dist_52w_low ┆ financial ┆ 0.076355 ┆ 0.016439 ┆ … ┆ 0.000003 ┆ 0.000178 ┆ true ┆ 16.152049 │\n",
|
||
"│ vol_252d ┆ financial ┆ 0.058948 ┆ 0.020559 ┆ … ┆ 0.004159 ┆ 0.068834 ┆ false ┆ 10.401608 │\n",
|
||
"│ natr_14 ┆ financial ┆ 0.05722 ┆ 0.01959 ┆ … ┆ 0.003506 ┆ 0.068834 ┆ false ┆ 10.388454 │\n",
|
||
"│ vol_126d ┆ financial ┆ 0.056434 ┆ 0.020275 ┆ … ┆ 0.005399 ┆ 0.068834 ┆ false ┆ 10.029753 │\n",
|
||
"│ vol_63d ┆ financial ┆ 0.05091 ┆ 0.019977 ┆ … ┆ 0.01085 ┆ 0.078944 ┆ false ┆ 9.110108 │\n",
|
||
"│ … ┆ … ┆ … ┆ … ┆ … ┆ … ┆ … ┆ … ┆ … │\n",
|
||
"│ ret_252d ┆ financial ┆ 0.040336 ┆ 0.018163 ┆ … ┆ 0.026411 ┆ 0.112247 ┆ false ┆ 7.997445 │\n",
|
||
"│ sma_ratio_200 ┆ financial ┆ 0.033548 ┆ 0.017222 ┆ … ┆ 0.051477 ┆ 0.187522 ┆ false ┆ 6.887362 │\n",
|
||
"│ mom_accel_sho ┆ financial ┆ -0.026099 ┆ 0.01662 ┆ … ┆ 0.116408 ┆ 0.395788 ┆ false ┆ -5.39076 │\n",
|
||
"│ rt ┆ ┆ ┆ ┆ ┆ ┆ ┆ ┆ │\n",
|
||
"│ skip_recent_6 ┆ financial ┆ 0.024087 ┆ 0.017551 ┆ … ┆ 0.169986 ┆ 0.471106 ┆ false ┆ 4.880859 │\n",
|
||
"│ _1 ┆ ┆ ┆ ┆ ┆ ┆ ┆ ┆ │\n",
|
||
"│ ret_126d ┆ financial ┆ 0.023384 ┆ 0.017382 ┆ … ┆ 0.178597 ┆ 0.471106 ┆ false ┆ 4.765151 │\n",
|
||
"└───────────────┴───────────┴───────────┴──────────┴───┴──────────┴──────────┴─────────┴───────────┘\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Top features table\n",
|
||
"print(\"\\nTop 15 features by absolute IC:\")\n",
|
||
"print(eval_summary.head(15))"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 15,
|
||
"id": "d42c4792",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:02.285202Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:02.285045Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:04.004496Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:04.003086Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"application/vnd.plotly.v1+json": {
|
||
"config": {
|
||
"plotlyServerURL": "https://plot.ly"
|
||
},
|
||
"data": [
|
||
{
|
||
"marker": {
|
||
"color": [
|
||
"#2ecc71",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6",
|
||
"#95a5a6"
|
||
]
|
||
},
|
||
"showlegend": false,
|
||
"text": [
|
||
"t=4.6",
|
||
"t=2.9",
|
||
"t=2.9",
|
||
"t=2.8",
|
||
"t=2.5",
|
||
"t=2.6",
|
||
"t=2.5",
|
||
"t=-2.5",
|
||
"t=2.3",
|
||
"t=2.3",
|
||
"t=2.2",
|
||
"t=1.9",
|
||
"t=-1.6",
|
||
"t=1.4",
|
||
"t=1.3",
|
||
"t=1.3",
|
||
"t=-1.3",
|
||
"t=1.2",
|
||
"t=-1.2",
|
||
"t=-1.0",
|
||
"t=1.0",
|
||
"t=-1.2",
|
||
"t=-1.3",
|
||
"t=-1.4",
|
||
"t=-1.2"
|
||
],
|
||
"textposition": "outside",
|
||
"type": "bar",
|
||
"x": [
|
||
"dist_52w_low",
|
||
"vol_252d",
|
||
"natr_14",
|
||
"vol_126d",
|
||
"vol_63d",
|
||
"vol_21d",
|
||
"garch_cond_vol",
|
||
"mom_accel_long",
|
||
"skip_recent_12_1",
|
||
"ret_189d",
|
||
"ret_252d",
|
||
"sma_ratio_200",
|
||
"mom_accel_short",
|
||
"skip_recent_6_1",
|
||
"ret_126d",
|
||
"ret_126d_rank",
|
||
"max_dd_63d",
|
||
"ret_63d",
|
||
"mom_accel_medium",
|
||
"max_dd_126d",
|
||
"sharpe_189d",
|
||
"vol_ratio_medium",
|
||
"bb_pctb_20",
|
||
"obv_zscore_63d",
|
||
"cci_21"
|
||
],
|
||
"xaxis": "x",
|
||
"y": [
|
||
0.07635483496346775,
|
||
0.05894811022065204,
|
||
0.05721956547277099,
|
||
0.056434302982184055,
|
||
0.05090995251982655,
|
||
0.050691280996018585,
|
||
0.04962564936381018,
|
||
-0.04325755427324245,
|
||
0.04178482635486817,
|
||
0.04110730011268391,
|
||
0.040335762919327035,
|
||
0.033548199627978344,
|
||
-0.02609920732671317,
|
||
0.02408747298706548,
|
||
0.02338393916006296,
|
||
0.02338393916006296,
|
||
-0.020294611560483112,
|
||
0.02014832251655958,
|
||
-0.020024239111810494,
|
||
-0.016060268548543852,
|
||
0.01602684766798153,
|
||
-0.015891984416866452,
|
||
-0.014715153790111598,
|
||
-0.014515752367914404,
|
||
-0.01398576327232167
|
||
],
|
||
"yaxis": "y"
|
||
},
|
||
{
|
||
"marker": {
|
||
"color": [
|
||
"#2ecc71",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c",
|
||
"#e74c3c"
|
||
],
|
||
"size": 7
|
||
},
|
||
"mode": "markers",
|
||
"showlegend": false,
|
||
"text": [
|
||
"dist_52w_low",
|
||
"vol_252d",
|
||
"natr_14",
|
||
"vol_126d",
|
||
"vol_63d",
|
||
"vol_21d",
|
||
"garch_cond_vol",
|
||
"mom_accel_long",
|
||
"skip_recent_12_1",
|
||
"ret_189d",
|
||
"ret_252d",
|
||
"sma_ratio_200",
|
||
"mom_accel_short",
|
||
"skip_recent_6_1",
|
||
"ret_126d",
|
||
"ret_126d_rank",
|
||
"max_dd_63d",
|
||
"ret_63d",
|
||
"mom_accel_medium",
|
||
"max_dd_126d",
|
||
"sharpe_189d",
|
||
"vol_ratio_medium",
|
||
"bb_pctb_20",
|
||
"obv_zscore_63d",
|
||
"cci_21",
|
||
"sharpe_252d",
|
||
"cci_14",
|
||
"sharpe_5d",
|
||
"vol_ratio_63d",
|
||
"ret_5d",
|
||
"vol_ratio_21d",
|
||
"sharpe_10d",
|
||
"rsi_7",
|
||
"sharpe_21d",
|
||
"stoch_k",
|
||
"sharpe_63d",
|
||
"rsi_14",
|
||
"hurst_100",
|
||
"adx_14",
|
||
"vol_ratio_short",
|
||
"dist_52w_high",
|
||
"ret_10d",
|
||
"ret_42d",
|
||
"macd_line",
|
||
"sma_ratio_50",
|
||
"sharpe_126d",
|
||
"sharpe_126d_rank",
|
||
"ret_21d",
|
||
"sharpe_42d",
|
||
"ema_ratio_26",
|
||
"chop_14"
|
||
],
|
||
"type": "scatter",
|
||
"x": [
|
||
16.152049300038176,
|
||
10.4016075507356,
|
||
10.388454162685514,
|
||
10.029752836301446,
|
||
9.110108471151076,
|
||
9.232100519933594,
|
||
8.985858205258696,
|
||
-9.031261408364609,
|
||
8.218590383587179,
|
||
8.220978593170363,
|
||
7.997444789955963,
|
||
6.887362192807514,
|
||
-5.390760069296785,
|
||
4.880859086460087,
|
||
4.765151116013638,
|
||
4.765151116013638,
|
||
-4.253847606584472,
|
||
4.160837207788063,
|
||
-4.1065294803346735,
|
||
-3.335857730998786,
|
||
3.4409608081391028,
|
||
-4.178142840935981,
|
||
-3.544083548045368,
|
||
-4.4289020641172065,
|
||
-3.369633808832531,
|
||
2.667595782857911,
|
||
-2.851685560525629,
|
||
-2.627904830616295,
|
||
4.685548712968539,
|
||
-2.123022458926743,
|
||
4.503854002005327,
|
||
-2.3656876386894536,
|
||
-2.2470797325178724,
|
||
-2.0358756479022198,
|
||
-1.9609186885402357,
|
||
1.7023350863586446,
|
||
-1.8104989769452282,
|
||
2.210259353934044,
|
||
-2.0217830995934047,
|
||
1.8968433729235636,
|
||
-1.2454355612086927,
|
||
-1.1830453071757785,
|
||
1.0167630917702568,
|
||
0.912949865812349,
|
||
0.5764095782358744,
|
||
0.586641688037098,
|
||
0.586641688037098,
|
||
-0.5431032988927165,
|
||
-0.5696204916308858,
|
||
0.11602668864737901,
|
||
0.0014786410254454996
|
||
],
|
||
"xaxis": "x2",
|
||
"y": [
|
||
4.644799243979996,
|
||
2.8672168229967703,
|
||
2.920918435289692,
|
||
2.7835057355409565,
|
||
2.5484905288863264,
|
||
2.6172624246564764,
|
||
2.5430734069348113,
|
||
-2.5019799111375485,
|
||
2.288662752088306,
|
||
2.2884594990169567,
|
||
2.2208066174556094,
|
||
1.9479762460285945,
|
||
-1.5703160371442408,
|
||
1.372458206342902,
|
||
1.3452848174210672,
|
||
1.3452848174210672,
|
||
-1.279006873146611,
|
||
1.1934812248147129,
|
||
-1.1747040781747424,
|
||
-0.9738188946252686,
|
||
0.9561759721124757,
|
||
-1.179690016521521,
|
||
-1.2885787414913032,
|
||
-1.3708409034763038,
|
||
-1.2087522043467198,
|
||
0.7226080456318031,
|
||
-1.1440855854213996,
|
||
-1.2543378930215137,
|
||
2.1174858835143358,
|
||
-1.0454372345102703,
|
||
2.423088919725733,
|
||
-0.8693722924491324,
|
||
-0.8277462682718482,
|
||
-0.6478307156265332,
|
||
-0.7548104791275595,
|
||
0.4829221193665501,
|
||
-0.5842713049070509,
|
||
0.6465805721307745,
|
||
-0.5788476286699612,
|
||
0.6130909085822979,
|
||
-0.35585342665912095,
|
||
-0.44218004949621936,
|
||
0.30491647319187715,
|
||
0.2705582651653039,
|
||
0.17513573998324436,
|
||
0.1657293406698883,
|
||
0.1657293406698883,
|
||
-0.17530775720162634,
|
||
-0.1658741178236516,
|
||
0.038394176596357445,
|
||
0.0005859349645728251
|
||
],
|
||
"yaxis": "y2"
|
||
},
|
||
{
|
||
"line": {
|
||
"color": "gray",
|
||
"dash": "dash"
|
||
},
|
||
"mode": "lines",
|
||
"showlegend": false,
|
||
"type": "scatter",
|
||
"x": [
|
||
-17.767254230041996,
|
||
17.767254230041996
|
||
],
|
||
"xaxis": "x2",
|
||
"y": [
|
||
-17.767254230041996,
|
||
17.767254230041996
|
||
],
|
||
"yaxis": "y2"
|
||
}
|
||
],
|
||
"layout": {
|
||
"annotations": [
|
||
{
|
||
"font": {
|
||
"size": 16
|
||
},
|
||
"showarrow": false,
|
||
"text": "Top 25 Features by |IC| (green = FDR-sig)",
|
||
"x": 0.22,
|
||
"xanchor": "center",
|
||
"xref": "paper",
|
||
"y": 1.0,
|
||
"yanchor": "bottom",
|
||
"yref": "paper"
|
||
},
|
||
{
|
||
"font": {
|
||
"size": 16
|
||
},
|
||
"showarrow": false,
|
||
"text": "HAC vs Naive t-statistics",
|
||
"x": 0.78,
|
||
"xanchor": "center",
|
||
"xref": "paper",
|
||
"y": 1.0,
|
||
"yanchor": "bottom",
|
||
"yref": "paper"
|
||
}
|
||
],
|
||
"height": 450,
|
||
"template": {
|
||
"data": {
|
||
"bar": [
|
||
{
|
||
"error_x": {
|
||
"color": "#2a3f5f"
|
||
},
|
||
"error_y": {
|
||
"color": "#2a3f5f"
|
||
},
|
||
"marker": {
|
||
"line": {
|
||
"color": "white",
|
||
"width": 0.5
|
||
},
|
||
"pattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
}
|
||
},
|
||
"type": "bar"
|
||
}
|
||
],
|
||
"barpolar": [
|
||
{
|
||
"marker": {
|
||
"line": {
|
||
"color": "white",
|
||
"width": 0.5
|
||
},
|
||
"pattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
}
|
||
},
|
||
"type": "barpolar"
|
||
}
|
||
],
|
||
"carpet": [
|
||
{
|
||
"aaxis": {
|
||
"endlinecolor": "#2a3f5f",
|
||
"gridcolor": "#C8D4E3",
|
||
"linecolor": "#C8D4E3",
|
||
"minorgridcolor": "#C8D4E3",
|
||
"startlinecolor": "#2a3f5f"
|
||
},
|
||
"baxis": {
|
||
"endlinecolor": "#2a3f5f",
|
||
"gridcolor": "#C8D4E3",
|
||
"linecolor": "#C8D4E3",
|
||
"minorgridcolor": "#C8D4E3",
|
||
"startlinecolor": "#2a3f5f"
|
||
},
|
||
"type": "carpet"
|
||
}
|
||
],
|
||
"choropleth": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"type": "choropleth"
|
||
}
|
||
],
|
||
"contour": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "contour"
|
||
}
|
||
],
|
||
"contourcarpet": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"type": "contourcarpet"
|
||
}
|
||
],
|
||
"heatmap": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "heatmap"
|
||
}
|
||
],
|
||
"histogram": [
|
||
{
|
||
"marker": {
|
||
"pattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
}
|
||
},
|
||
"type": "histogram"
|
||
}
|
||
],
|
||
"histogram2d": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "histogram2d"
|
||
}
|
||
],
|
||
"histogram2dcontour": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "histogram2dcontour"
|
||
}
|
||
],
|
||
"mesh3d": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"type": "mesh3d"
|
||
}
|
||
],
|
||
"parcoords": [
|
||
{
|
||
"line": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "parcoords"
|
||
}
|
||
],
|
||
"pie": [
|
||
{
|
||
"automargin": true,
|
||
"type": "pie"
|
||
}
|
||
],
|
||
"scatter": [
|
||
{
|
||
"fillpattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
},
|
||
"type": "scatter"
|
||
}
|
||
],
|
||
"scatter3d": [
|
||
{
|
||
"line": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatter3d"
|
||
}
|
||
],
|
||
"scattercarpet": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattercarpet"
|
||
}
|
||
],
|
||
"scattergeo": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattergeo"
|
||
}
|
||
],
|
||
"scattergl": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattergl"
|
||
}
|
||
],
|
||
"scattermap": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattermap"
|
||
}
|
||
],
|
||
"scattermapbox": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattermapbox"
|
||
}
|
||
],
|
||
"scatterpolar": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatterpolar"
|
||
}
|
||
],
|
||
"scatterpolargl": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatterpolargl"
|
||
}
|
||
],
|
||
"scatterternary": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatterternary"
|
||
}
|
||
],
|
||
"surface": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "surface"
|
||
}
|
||
],
|
||
"table": [
|
||
{
|
||
"cells": {
|
||
"fill": {
|
||
"color": "#EBF0F8"
|
||
},
|
||
"line": {
|
||
"color": "white"
|
||
}
|
||
},
|
||
"header": {
|
||
"fill": {
|
||
"color": "#C8D4E3"
|
||
},
|
||
"line": {
|
||
"color": "white"
|
||
}
|
||
},
|
||
"type": "table"
|
||
}
|
||
]
|
||
},
|
||
"layout": {
|
||
"annotationdefaults": {
|
||
"arrowcolor": "#2a3f5f",
|
||
"arrowhead": 0,
|
||
"arrowwidth": 1
|
||
},
|
||
"autotypenumbers": "strict",
|
||
"coloraxis": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"colorscale": {
|
||
"diverging": [
|
||
[
|
||
0,
|
||
"#8e0152"
|
||
],
|
||
[
|
||
0.1,
|
||
"#c51b7d"
|
||
],
|
||
[
|
||
0.2,
|
||
"#de77ae"
|
||
],
|
||
[
|
||
0.3,
|
||
"#f1b6da"
|
||
],
|
||
[
|
||
0.4,
|
||
"#fde0ef"
|
||
],
|
||
[
|
||
0.5,
|
||
"#f7f7f7"
|
||
],
|
||
[
|
||
0.6,
|
||
"#e6f5d0"
|
||
],
|
||
[
|
||
0.7,
|
||
"#b8e186"
|
||
],
|
||
[
|
||
0.8,
|
||
"#7fbc41"
|
||
],
|
||
[
|
||
0.9,
|
||
"#4d9221"
|
||
],
|
||
[
|
||
1,
|
||
"#276419"
|
||
]
|
||
],
|
||
"sequential": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"sequentialminus": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
]
|
||
},
|
||
"colorway": [
|
||
"#636efa",
|
||
"#EF553B",
|
||
"#00cc96",
|
||
"#ab63fa",
|
||
"#FFA15A",
|
||
"#19d3f3",
|
||
"#FF6692",
|
||
"#B6E880",
|
||
"#FF97FF",
|
||
"#FECB52"
|
||
],
|
||
"font": {
|
||
"color": "#2a3f5f"
|
||
},
|
||
"geo": {
|
||
"bgcolor": "white",
|
||
"lakecolor": "white",
|
||
"landcolor": "white",
|
||
"showlakes": true,
|
||
"showland": true,
|
||
"subunitcolor": "#C8D4E3"
|
||
},
|
||
"hoverlabel": {
|
||
"align": "left"
|
||
},
|
||
"hovermode": "closest",
|
||
"mapbox": {
|
||
"style": "light"
|
||
},
|
||
"paper_bgcolor": "white",
|
||
"plot_bgcolor": "white",
|
||
"polar": {
|
||
"angularaxis": {
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": ""
|
||
},
|
||
"bgcolor": "white",
|
||
"radialaxis": {
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"scene": {
|
||
"xaxis": {
|
||
"backgroundcolor": "white",
|
||
"gridcolor": "#DFE8F3",
|
||
"gridwidth": 2,
|
||
"linecolor": "#EBF0F8",
|
||
"showbackground": true,
|
||
"ticks": "",
|
||
"zerolinecolor": "#EBF0F8"
|
||
},
|
||
"yaxis": {
|
||
"backgroundcolor": "white",
|
||
"gridcolor": "#DFE8F3",
|
||
"gridwidth": 2,
|
||
"linecolor": "#EBF0F8",
|
||
"showbackground": true,
|
||
"ticks": "",
|
||
"zerolinecolor": "#EBF0F8"
|
||
},
|
||
"zaxis": {
|
||
"backgroundcolor": "white",
|
||
"gridcolor": "#DFE8F3",
|
||
"gridwidth": 2,
|
||
"linecolor": "#EBF0F8",
|
||
"showbackground": true,
|
||
"ticks": "",
|
||
"zerolinecolor": "#EBF0F8"
|
||
}
|
||
},
|
||
"shapedefaults": {
|
||
"line": {
|
||
"color": "#2a3f5f"
|
||
}
|
||
},
|
||
"ternary": {
|
||
"aaxis": {
|
||
"gridcolor": "#DFE8F3",
|
||
"linecolor": "#A2B1C6",
|
||
"ticks": ""
|
||
},
|
||
"baxis": {
|
||
"gridcolor": "#DFE8F3",
|
||
"linecolor": "#A2B1C6",
|
||
"ticks": ""
|
||
},
|
||
"bgcolor": "white",
|
||
"caxis": {
|
||
"gridcolor": "#DFE8F3",
|
||
"linecolor": "#A2B1C6",
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"title": {
|
||
"x": 0.05
|
||
},
|
||
"xaxis": {
|
||
"automargin": true,
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": "",
|
||
"title": {
|
||
"standoff": 15
|
||
},
|
||
"zerolinecolor": "#EBF0F8",
|
||
"zerolinewidth": 2
|
||
},
|
||
"yaxis": {
|
||
"automargin": true,
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": "",
|
||
"title": {
|
||
"standoff": 15
|
||
},
|
||
"zerolinecolor": "#EBF0F8",
|
||
"zerolinewidth": 2
|
||
}
|
||
}
|
||
},
|
||
"width": 1100,
|
||
"xaxis": {
|
||
"anchor": "y",
|
||
"domain": [
|
||
0.0,
|
||
0.44
|
||
],
|
||
"tickangle": -45
|
||
},
|
||
"xaxis2": {
|
||
"anchor": "y2",
|
||
"domain": [
|
||
0.56,
|
||
1.0
|
||
],
|
||
"title": {
|
||
"text": "Naive t"
|
||
}
|
||
},
|
||
"yaxis": {
|
||
"anchor": "x",
|
||
"domain": [
|
||
0.0,
|
||
1.0
|
||
]
|
||
},
|
||
"yaxis2": {
|
||
"anchor": "x2",
|
||
"domain": [
|
||
0.0,
|
||
1.0
|
||
],
|
||
"title": {
|
||
"text": "HAC t"
|
||
}
|
||
}
|
||
}
|
||
},
|
||
"image/png": "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"
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"top_n = min(25, len(eval_summary))\n",
|
||
"top = eval_summary.head(top_n)\n",
|
||
"\n",
|
||
"fig = make_subplots(\n",
|
||
" rows=1,\n",
|
||
" cols=2,\n",
|
||
" subplot_titles=[\n",
|
||
" f\"Top {top_n} Features by |IC| (green = FDR-sig)\",\n",
|
||
" \"HAC vs Naive t-statistics\",\n",
|
||
" ],\n",
|
||
" horizontal_spacing=0.12,\n",
|
||
")\n",
|
||
"\n",
|
||
"# Panel 1: IC bar chart\n",
|
||
"colors = [\"#2ecc71\" if s else \"#95a5a6\" for s in top[\"fdr_sig\"].to_list()]\n",
|
||
"fig.add_trace(\n",
|
||
" go.Bar(\n",
|
||
" x=top[\"feature\"].to_list(),\n",
|
||
" y=top[\"ic_mean\"].to_list(),\n",
|
||
" marker_color=colors,\n",
|
||
" text=[f\"t={t:.1f}\" for t in top[\"hac_t\"].to_list()],\n",
|
||
" textposition=\"outside\",\n",
|
||
" showlegend=False,\n",
|
||
" ),\n",
|
||
" row=1,\n",
|
||
" col=1,\n",
|
||
")\n",
|
||
"\n",
|
||
"# Panel 2: HAC scatter\n",
|
||
"fig.add_trace(\n",
|
||
" go.Scatter(\n",
|
||
" x=eval_summary[\"naive_t\"].to_list(),\n",
|
||
" y=eval_summary[\"hac_t\"].to_list(),\n",
|
||
" mode=\"markers\",\n",
|
||
" marker=dict(\n",
|
||
" color=[\"#2ecc71\" if s else \"#e74c3c\" for s in eval_summary[\"fdr_sig\"].to_list()],\n",
|
||
" size=7,\n",
|
||
" ),\n",
|
||
" text=eval_summary[\"feature\"].to_list(),\n",
|
||
" showlegend=False,\n",
|
||
" ),\n",
|
||
" row=1,\n",
|
||
" col=2,\n",
|
||
")\n",
|
||
"max_t = (\n",
|
||
" max(\n",
|
||
" eval_summary[\"naive_t\"].cast(pl.Float64, strict=False).abs().max() or 1.0,\n",
|
||
" eval_summary[\"hac_t\"].cast(pl.Float64, strict=False).abs().max() or 1.0,\n",
|
||
" )\n",
|
||
" * 1.1\n",
|
||
")\n",
|
||
"fig.add_trace(\n",
|
||
" go.Scatter(\n",
|
||
" x=[-max_t, max_t],\n",
|
||
" y=[-max_t, max_t],\n",
|
||
" mode=\"lines\",\n",
|
||
" line=dict(dash=\"dash\", color=\"gray\"),\n",
|
||
" showlegend=False,\n",
|
||
" ),\n",
|
||
" row=1,\n",
|
||
" col=2,\n",
|
||
")\n",
|
||
"\n",
|
||
"fig.update_layout(template=\"plotly_white\", height=450, width=1100)\n",
|
||
"fig.update_xaxes(tickangle=-45, row=1, col=1)\n",
|
||
"fig.update_xaxes(title_text=\"Naive t\", row=1, col=2)\n",
|
||
"fig.update_yaxes(title_text=\"HAC t\", row=1, col=2)\n",
|
||
"fig.show()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "6fe36d9b",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"**Interpretation**: Points below the diagonal in the right panel show where\n",
|
||
"HAC adjustment deflates naive t-statistics. With overlapping 21-day returns,\n",
|
||
"daily IC values are autocorrelated and naive standard errors understate\n",
|
||
"uncertainty. The green markers survive both HAC and FDR correction — these\n",
|
||
"features have genuine standalone predictive power."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "47346f68",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"## 4. Shape Diagnostics\n",
|
||
"\n",
|
||
"Quantile monotonicity analysis: does the label spread monotonically across\n",
|
||
"feature quintiles? A monotone relationship (Q1 < Q2 < ... < Q5 or reverse)\n",
|
||
"suggests a robust, exploitable signal. Non-monotone shapes may indicate\n",
|
||
"non-linear interactions or noise."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 16,
|
||
"id": "60e70a13",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:04.060411Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:04.060071Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:04.101499Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:04.100284Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Shape analysis for 1 features\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"N_QUANTILES = 5\n",
|
||
"top_features_for_shape = eval_summary.filter(pl.col(\"fdr_sig\").fill_null(False))[\n",
|
||
" \"feature\"\n",
|
||
"].to_list()[:15]\n",
|
||
"if not top_features_for_shape:\n",
|
||
" top_features_for_shape = eval_summary.head(10)[\"feature\"].to_list()\n",
|
||
"\n",
|
||
"monotonicity_scores = {}\n",
|
||
"quantile_spreads = {}\n",
|
||
"\n",
|
||
"for feat in top_features_for_shape:\n",
|
||
" valid = eval_panel.select([feat, label_col]).drop_nulls()\n",
|
||
" if len(valid) < N_QUANTILES * 20:\n",
|
||
" continue\n",
|
||
"\n",
|
||
" valid = valid.with_columns(\n",
|
||
" pl.col(feat)\n",
|
||
" .qcut(N_QUANTILES, labels=[f\"Q{i + 1}\" for i in range(N_QUANTILES)])\n",
|
||
" .alias(\"quantile\")\n",
|
||
" )\n",
|
||
" q_means = valid.group_by(\"quantile\").agg(pl.col(label_col).mean()).sort(\"quantile\")\n",
|
||
" means = q_means[label_col].to_list()\n",
|
||
" spread = means[-1] - means[0]\n",
|
||
" quantile_spreads[feat] = {\"q_means\": means, \"spread\": spread}\n",
|
||
"\n",
|
||
" mono_corr, _ = spearmanr(range(len(means)), means)\n",
|
||
" monotonicity_scores[feat] = float(mono_corr)\n",
|
||
"\n",
|
||
"print(f\"Shape analysis for {len(quantile_spreads)} features\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 17,
|
||
"id": "9b2e7b93",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:04.122580Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:04.122128Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:04.270326Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:04.269519Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"application/vnd.plotly.v1+json": {
|
||
"config": {
|
||
"plotlyServerURL": "https://plot.ly"
|
||
},
|
||
"data": [
|
||
{
|
||
"marker": {
|
||
"color": [
|
||
"#e74c3c",
|
||
"#f39c12",
|
||
"#95a5a6",
|
||
"#3498db",
|
||
"#2ecc71"
|
||
]
|
||
},
|
||
"showlegend": false,
|
||
"text": [
|
||
"0.0040",
|
||
"0.0039",
|
||
"0.0052",
|
||
"0.0083",
|
||
"0.0133"
|
||
],
|
||
"textposition": "outside",
|
||
"type": "bar",
|
||
"x": [
|
||
"Q1",
|
||
"Q2",
|
||
"Q3",
|
||
"Q4",
|
||
"Q5"
|
||
],
|
||
"xaxis": "x",
|
||
"y": [
|
||
0.00395786106430973,
|
||
0.0038895872583125635,
|
||
0.005215263716603622,
|
||
0.008298399870340579,
|
||
0.013263235621217367
|
||
],
|
||
"yaxis": "y"
|
||
}
|
||
],
|
||
"layout": {
|
||
"annotations": [
|
||
{
|
||
"font": {
|
||
"size": 16
|
||
},
|
||
"showarrow": false,
|
||
"text": "dist_52w_low",
|
||
"x": 0.14444444444444446,
|
||
"xanchor": "center",
|
||
"xref": "paper",
|
||
"y": 1.0,
|
||
"yanchor": "bottom",
|
||
"yref": "paper"
|
||
}
|
||
],
|
||
"height": 250,
|
||
"template": {
|
||
"data": {
|
||
"bar": [
|
||
{
|
||
"error_x": {
|
||
"color": "#2a3f5f"
|
||
},
|
||
"error_y": {
|
||
"color": "#2a3f5f"
|
||
},
|
||
"marker": {
|
||
"line": {
|
||
"color": "white",
|
||
"width": 0.5
|
||
},
|
||
"pattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
}
|
||
},
|
||
"type": "bar"
|
||
}
|
||
],
|
||
"barpolar": [
|
||
{
|
||
"marker": {
|
||
"line": {
|
||
"color": "white",
|
||
"width": 0.5
|
||
},
|
||
"pattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
}
|
||
},
|
||
"type": "barpolar"
|
||
}
|
||
],
|
||
"carpet": [
|
||
{
|
||
"aaxis": {
|
||
"endlinecolor": "#2a3f5f",
|
||
"gridcolor": "#C8D4E3",
|
||
"linecolor": "#C8D4E3",
|
||
"minorgridcolor": "#C8D4E3",
|
||
"startlinecolor": "#2a3f5f"
|
||
},
|
||
"baxis": {
|
||
"endlinecolor": "#2a3f5f",
|
||
"gridcolor": "#C8D4E3",
|
||
"linecolor": "#C8D4E3",
|
||
"minorgridcolor": "#C8D4E3",
|
||
"startlinecolor": "#2a3f5f"
|
||
},
|
||
"type": "carpet"
|
||
}
|
||
],
|
||
"choropleth": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"type": "choropleth"
|
||
}
|
||
],
|
||
"contour": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "contour"
|
||
}
|
||
],
|
||
"contourcarpet": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"type": "contourcarpet"
|
||
}
|
||
],
|
||
"heatmap": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "heatmap"
|
||
}
|
||
],
|
||
"histogram": [
|
||
{
|
||
"marker": {
|
||
"pattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
}
|
||
},
|
||
"type": "histogram"
|
||
}
|
||
],
|
||
"histogram2d": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "histogram2d"
|
||
}
|
||
],
|
||
"histogram2dcontour": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "histogram2dcontour"
|
||
}
|
||
],
|
||
"mesh3d": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"type": "mesh3d"
|
||
}
|
||
],
|
||
"parcoords": [
|
||
{
|
||
"line": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "parcoords"
|
||
}
|
||
],
|
||
"pie": [
|
||
{
|
||
"automargin": true,
|
||
"type": "pie"
|
||
}
|
||
],
|
||
"scatter": [
|
||
{
|
||
"fillpattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
},
|
||
"type": "scatter"
|
||
}
|
||
],
|
||
"scatter3d": [
|
||
{
|
||
"line": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatter3d"
|
||
}
|
||
],
|
||
"scattercarpet": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattercarpet"
|
||
}
|
||
],
|
||
"scattergeo": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattergeo"
|
||
}
|
||
],
|
||
"scattergl": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattergl"
|
||
}
|
||
],
|
||
"scattermap": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattermap"
|
||
}
|
||
],
|
||
"scattermapbox": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattermapbox"
|
||
}
|
||
],
|
||
"scatterpolar": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatterpolar"
|
||
}
|
||
],
|
||
"scatterpolargl": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatterpolargl"
|
||
}
|
||
],
|
||
"scatterternary": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatterternary"
|
||
}
|
||
],
|
||
"surface": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "surface"
|
||
}
|
||
],
|
||
"table": [
|
||
{
|
||
"cells": {
|
||
"fill": {
|
||
"color": "#EBF0F8"
|
||
},
|
||
"line": {
|
||
"color": "white"
|
||
}
|
||
},
|
||
"header": {
|
||
"fill": {
|
||
"color": "#C8D4E3"
|
||
},
|
||
"line": {
|
||
"color": "white"
|
||
}
|
||
},
|
||
"type": "table"
|
||
}
|
||
]
|
||
},
|
||
"layout": {
|
||
"annotationdefaults": {
|
||
"arrowcolor": "#2a3f5f",
|
||
"arrowhead": 0,
|
||
"arrowwidth": 1
|
||
},
|
||
"autotypenumbers": "strict",
|
||
"coloraxis": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"colorscale": {
|
||
"diverging": [
|
||
[
|
||
0,
|
||
"#8e0152"
|
||
],
|
||
[
|
||
0.1,
|
||
"#c51b7d"
|
||
],
|
||
[
|
||
0.2,
|
||
"#de77ae"
|
||
],
|
||
[
|
||
0.3,
|
||
"#f1b6da"
|
||
],
|
||
[
|
||
0.4,
|
||
"#fde0ef"
|
||
],
|
||
[
|
||
0.5,
|
||
"#f7f7f7"
|
||
],
|
||
[
|
||
0.6,
|
||
"#e6f5d0"
|
||
],
|
||
[
|
||
0.7,
|
||
"#b8e186"
|
||
],
|
||
[
|
||
0.8,
|
||
"#7fbc41"
|
||
],
|
||
[
|
||
0.9,
|
||
"#4d9221"
|
||
],
|
||
[
|
||
1,
|
||
"#276419"
|
||
]
|
||
],
|
||
"sequential": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"sequentialminus": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
]
|
||
},
|
||
"colorway": [
|
||
"#636efa",
|
||
"#EF553B",
|
||
"#00cc96",
|
||
"#ab63fa",
|
||
"#FFA15A",
|
||
"#19d3f3",
|
||
"#FF6692",
|
||
"#B6E880",
|
||
"#FF97FF",
|
||
"#FECB52"
|
||
],
|
||
"font": {
|
||
"color": "#2a3f5f"
|
||
},
|
||
"geo": {
|
||
"bgcolor": "white",
|
||
"lakecolor": "white",
|
||
"landcolor": "white",
|
||
"showlakes": true,
|
||
"showland": true,
|
||
"subunitcolor": "#C8D4E3"
|
||
},
|
||
"hoverlabel": {
|
||
"align": "left"
|
||
},
|
||
"hovermode": "closest",
|
||
"mapbox": {
|
||
"style": "light"
|
||
},
|
||
"paper_bgcolor": "white",
|
||
"plot_bgcolor": "white",
|
||
"polar": {
|
||
"angularaxis": {
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": ""
|
||
},
|
||
"bgcolor": "white",
|
||
"radialaxis": {
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"scene": {
|
||
"xaxis": {
|
||
"backgroundcolor": "white",
|
||
"gridcolor": "#DFE8F3",
|
||
"gridwidth": 2,
|
||
"linecolor": "#EBF0F8",
|
||
"showbackground": true,
|
||
"ticks": "",
|
||
"zerolinecolor": "#EBF0F8"
|
||
},
|
||
"yaxis": {
|
||
"backgroundcolor": "white",
|
||
"gridcolor": "#DFE8F3",
|
||
"gridwidth": 2,
|
||
"linecolor": "#EBF0F8",
|
||
"showbackground": true,
|
||
"ticks": "",
|
||
"zerolinecolor": "#EBF0F8"
|
||
},
|
||
"zaxis": {
|
||
"backgroundcolor": "white",
|
||
"gridcolor": "#DFE8F3",
|
||
"gridwidth": 2,
|
||
"linecolor": "#EBF0F8",
|
||
"showbackground": true,
|
||
"ticks": "",
|
||
"zerolinecolor": "#EBF0F8"
|
||
}
|
||
},
|
||
"shapedefaults": {
|
||
"line": {
|
||
"color": "#2a3f5f"
|
||
}
|
||
},
|
||
"ternary": {
|
||
"aaxis": {
|
||
"gridcolor": "#DFE8F3",
|
||
"linecolor": "#A2B1C6",
|
||
"ticks": ""
|
||
},
|
||
"baxis": {
|
||
"gridcolor": "#DFE8F3",
|
||
"linecolor": "#A2B1C6",
|
||
"ticks": ""
|
||
},
|
||
"bgcolor": "white",
|
||
"caxis": {
|
||
"gridcolor": "#DFE8F3",
|
||
"linecolor": "#A2B1C6",
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"title": {
|
||
"x": 0.05
|
||
},
|
||
"xaxis": {
|
||
"automargin": true,
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": "",
|
||
"title": {
|
||
"standoff": 15
|
||
},
|
||
"zerolinecolor": "#EBF0F8",
|
||
"zerolinewidth": 2
|
||
},
|
||
"yaxis": {
|
||
"automargin": true,
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": "",
|
||
"title": {
|
||
"standoff": 15
|
||
},
|
||
"zerolinecolor": "#EBF0F8",
|
||
"zerolinewidth": 2
|
||
}
|
||
}
|
||
},
|
||
"title": {
|
||
"text": "Quantile Mean Returns (Top FDR-Significant Features)"
|
||
},
|
||
"width": 900,
|
||
"xaxis": {
|
||
"anchor": "y",
|
||
"domain": [
|
||
0.0,
|
||
0.2888888888888889
|
||
]
|
||
},
|
||
"xaxis2": {
|
||
"anchor": "y2",
|
||
"domain": [
|
||
0.35555555555555557,
|
||
0.6444444444444445
|
||
]
|
||
},
|
||
"xaxis3": {
|
||
"anchor": "y3",
|
||
"domain": [
|
||
0.7111111111111111,
|
||
1.0
|
||
]
|
||
},
|
||
"yaxis": {
|
||
"anchor": "x",
|
||
"domain": [
|
||
0.0,
|
||
1.0
|
||
]
|
||
},
|
||
"yaxis2": {
|
||
"anchor": "x2",
|
||
"domain": [
|
||
0.0,
|
||
1.0
|
||
]
|
||
},
|
||
"yaxis3": {
|
||
"anchor": "x3",
|
||
"domain": [
|
||
0.0,
|
||
1.0
|
||
]
|
||
}
|
||
}
|
||
},
|
||
"image/png": "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"
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"if quantile_spreads:\n",
|
||
" n_show = min(6, len(quantile_spreads))\n",
|
||
" feats_to_show = list(quantile_spreads.keys())[:n_show]\n",
|
||
" n_rows_fig = (n_show + 2) // 3\n",
|
||
" fig = make_subplots(rows=n_rows_fig, cols=3, subplot_titles=feats_to_show)\n",
|
||
" for idx, feat in enumerate(feats_to_show):\n",
|
||
" r, c = divmod(idx, 3)\n",
|
||
" q_means = quantile_spreads[feat][\"q_means\"]\n",
|
||
" mono = monotonicity_scores.get(feat, 0)\n",
|
||
" fig.add_trace(\n",
|
||
" go.Bar(\n",
|
||
" x=[f\"Q{i + 1}\" for i in range(len(q_means))],\n",
|
||
" y=q_means,\n",
|
||
" marker_color=[\n",
|
||
" \"#e74c3c\",\n",
|
||
" \"#f39c12\",\n",
|
||
" \"#95a5a6\",\n",
|
||
" \"#3498db\",\n",
|
||
" \"#2ecc71\",\n",
|
||
" ],\n",
|
||
" showlegend=False,\n",
|
||
" text=[f\"{m:.4f}\" for m in q_means],\n",
|
||
" textposition=\"outside\",\n",
|
||
" ),\n",
|
||
" row=r + 1,\n",
|
||
" col=c + 1,\n",
|
||
" )\n",
|
||
" fig.update_layout(\n",
|
||
" template=\"plotly_white\",\n",
|
||
" height=250 * n_rows_fig,\n",
|
||
" width=900,\n",
|
||
" title_text=\"Quantile Mean Returns (Top FDR-Significant Features)\",\n",
|
||
" )\n",
|
||
" fig.show()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "ab210a0f",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"**Interpretation**: Monotone quintile spreads (Q1 → Q5 increasing or\n",
|
||
"decreasing) confirm a robust relationship. Non-monotone patterns (e.g.,\n",
|
||
"U-shaped) suggest non-linear effects that may require interaction terms or\n",
|
||
"tree-based models to capture."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "399cab4a",
|
||
"metadata": {
|
||
"lines_to_next_cell": 2,
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"## 5. Redundancy & Feature Families\n",
|
||
"\n",
|
||
"We group features into interpretive families and compute family-level IC\n",
|
||
"aggregates. Highly correlated feature pairs (|corr| > 0.7) flag redundancy\n",
|
||
"that downstream modeling should address via clustering or selection."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 18,
|
||
"id": "6f132825",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:04.318642Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:04.318279Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:04.326621Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:04.325596Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"def assign_feature_family(feature_name: str) -> str:\n",
|
||
" \"\"\"Map feature name to family based on prefix.\"\"\"\n",
|
||
" FAMILY_MAP = [\n",
|
||
" ([\"sharpe_\", \"risk_adj\"], \"risk_adj_momentum\"),\n",
|
||
" ([\"skip_recent\", \"mom_\"], \"momentum\"),\n",
|
||
" ([\"ret_\"], \"momentum\"),\n",
|
||
" ([\"vol_ratio\"], \"vol_ratio\"),\n",
|
||
" ([\"vol_\"], \"volatility\"),\n",
|
||
" ([\"natr\", \"range_\", \"max_dd\"], \"volatility\"),\n",
|
||
" ([\"rsi\", \"macd\", \"adx\", \"cci\", \"stoch\", \"aroon\", \"bb_\"], \"technical\"),\n",
|
||
" ([\"sma_\", \"ema_\"], \"trend\"),\n",
|
||
" ([\"yield_curve\", \"spy_tlt\", \"regime\", \"chop\", \"hurst\"], \"regime\"),\n",
|
||
" ([\"rank_\"], \"cross_sectional\"),\n",
|
||
" ([\"obv\", \"turnover\", \"volume\"], \"volume\"),\n",
|
||
" ([\"pct_positive\", \"up_ratio\"], \"consistency\"),\n",
|
||
" ([\"dist_\"], \"distance\"),\n",
|
||
" ([\"corr_\"], \"correlation\"),\n",
|
||
" ]\n",
|
||
" for prefixes, family in FAMILY_MAP:\n",
|
||
" if any(p in feature_name.lower() for p in prefixes):\n",
|
||
" return family\n",
|
||
" return \"other\"\n",
|
||
"\n",
|
||
"\n",
|
||
"families = {feat: assign_feature_family(feat) for feat in all_feature_cols}\n",
|
||
"\n",
|
||
"# Override temporal features with specific families\n",
|
||
"for feat in temporal_cols:\n",
|
||
" if \"regime\" in feat.lower():\n",
|
||
" families[feat] = \"temporal_regime\"\n",
|
||
" elif \"ffd\" in feat.lower():\n",
|
||
" families[feat] = \"temporal_ffd\"\n",
|
||
" else:\n",
|
||
" families[feat] = \"temporal_other\""
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 19,
|
||
"id": "bcb9dbe1",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:04.340330Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:04.340124Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:07.673984Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:07.668918Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Feature pairs with |corr| > 0.7: 157\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Pairwise correlation on sampled dates\n",
|
||
"sample_step = max(1, n_dates // 200)\n",
|
||
"sample_dates = eval_panel[DATE_COL].unique().sort().to_list()[::sample_step]\n",
|
||
"corr_data = (\n",
|
||
" eval_panel.filter(pl.col(DATE_COL).is_in(sample_dates)).select(evaluable_features).to_pandas()\n",
|
||
")\n",
|
||
"corr_matrix = corr_data.corr(method=\"spearman\")\n",
|
||
"\n",
|
||
"high_corr_pairs = []\n",
|
||
"cols = corr_matrix.columns\n",
|
||
"for i in range(len(cols)):\n",
|
||
" for j in range(i + 1, len(cols)):\n",
|
||
" if abs(corr_matrix.iloc[i, j]) > 0.7:\n",
|
||
" high_corr_pairs.append((cols[i], cols[j], float(corr_matrix.iloc[i, j])))\n",
|
||
"\n",
|
||
"print(f\"Feature pairs with |corr| > 0.7: {len(high_corr_pairs)}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 20,
|
||
"id": "b438101b",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:07.702801Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:07.702314Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:07.721190Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:07.719396Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"shape: (10, 5)\n",
|
||
"┌───────────────────┬────────────┬────────────┬───────────┬───────────┐\n",
|
||
"│ family ┆ n_features ┆ avg_abs_ic ┆ avg_ic ┆ n_fdr_sig │\n",
|
||
"│ --- ┆ --- ┆ --- ┆ --- ┆ --- │\n",
|
||
"│ str ┆ i64 ┆ f64 ┆ f64 ┆ i64 │\n",
|
||
"╞═══════════════════╪════════════╪════════════╪═══════════╪═══════════╡\n",
|
||
"│ temporal_other ┆ 1 ┆ 0.049626 ┆ 0.049626 ┆ 0 │\n",
|
||
"│ volatility ┆ 7 ┆ 0.044365 ┆ 0.033978 ┆ 0 │\n",
|
||
"│ distance ┆ 2 ┆ 0.041219 ┆ 0.035135 ┆ 1 │\n",
|
||
"│ momentum ┆ 14 ┆ 0.023319 ┆ 0.007958 ┆ 0 │\n",
|
||
"│ volume ┆ 1 ┆ 0.014516 ┆ -0.014516 ┆ 0 │\n",
|
||
"│ trend ┆ 3 ┆ 0.012244 ┆ 0.012244 ┆ 0 │\n",
|
||
"│ vol_ratio ┆ 4 ┆ 0.010735 ┆ 0.002789 ┆ 0 │\n",
|
||
"│ technical ┆ 8 ┆ 0.009498 ┆ -0.008539 ┆ 0 │\n",
|
||
"│ risk_adj_momentum ┆ 9 ┆ 0.008073 ┆ 0.001112 ┆ 0 │\n",
|
||
"│ regime ┆ 2 ┆ 0.003418 ┆ 0.003418 ┆ 0 │\n",
|
||
"└───────────────────┴────────────┴────────────┴───────────┴───────────┘\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Family-level IC summary\n",
|
||
"family_ic = {}\n",
|
||
"fdr_sig_set = set(eval_summary.filter(pl.col(\"fdr_sig\").fill_null(False))[\"feature\"].to_list())\n",
|
||
"\n",
|
||
"for feat in ic_results:\n",
|
||
" fam = families.get(feat, \"other\")\n",
|
||
" family_ic.setdefault(fam, []).append(\n",
|
||
" {\n",
|
||
" \"feature\": feat,\n",
|
||
" \"ic\": ic_results[feat][\"mean_ic\"],\n",
|
||
" \"fdr_sig\": feat in fdr_sig_set,\n",
|
||
" }\n",
|
||
" )\n",
|
||
"\n",
|
||
"family_summary = {}\n",
|
||
"for fam, feats in sorted(family_ic.items()):\n",
|
||
" ics = [f[\"ic\"] for f in feats if f[\"ic\"] is not None]\n",
|
||
" n_sig = sum(1 for f in feats if f[\"fdr_sig\"])\n",
|
||
" family_summary[fam] = {\n",
|
||
" \"n_features\": len(feats),\n",
|
||
" \"avg_abs_ic\": float(np.mean([abs(ic) for ic in ics])) if ics else 0.0,\n",
|
||
" \"avg_ic\": float(np.mean(ics)) if ics else 0.0,\n",
|
||
" \"n_fdr_sig\": n_sig,\n",
|
||
" }\n",
|
||
"\n",
|
||
"if family_summary:\n",
|
||
" fam_df = pl.DataFrame([{\"family\": fam, **stats} for fam, stats in family_summary.items()]).sort(\n",
|
||
" \"avg_abs_ic\", descending=True\n",
|
||
" )\n",
|
||
"else:\n",
|
||
" fam_df = pl.DataFrame(\n",
|
||
" schema={\n",
|
||
" \"family\": pl.Utf8,\n",
|
||
" \"n_features\": pl.Int64,\n",
|
||
" \"avg_abs_ic\": pl.Float64,\n",
|
||
" \"avg_ic\": pl.Float64,\n",
|
||
" \"n_fdr_sig\": pl.Int64,\n",
|
||
" }\n",
|
||
" )\n",
|
||
"print(fam_df)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 21,
|
||
"id": "023b2853",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:07.743975Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:07.743428Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:08.099823Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:08.098752Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"application/vnd.plotly.v1+json": {
|
||
"config": {
|
||
"plotlyServerURL": "https://plot.ly"
|
||
},
|
||
"data": [
|
||
{
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"rgb(5,48,97)"
|
||
],
|
||
[
|
||
0.1,
|
||
"rgb(33,102,172)"
|
||
],
|
||
[
|
||
0.2,
|
||
"rgb(67,147,195)"
|
||
],
|
||
[
|
||
0.3,
|
||
"rgb(146,197,222)"
|
||
],
|
||
[
|
||
0.4,
|
||
"rgb(209,229,240)"
|
||
],
|
||
[
|
||
0.5,
|
||
"rgb(247,247,247)"
|
||
],
|
||
[
|
||
0.6,
|
||
"rgb(253,219,199)"
|
||
],
|
||
[
|
||
0.7,
|
||
"rgb(244,165,130)"
|
||
],
|
||
[
|
||
0.8,
|
||
"rgb(214,96,77)"
|
||
],
|
||
[
|
||
0.9,
|
||
"rgb(178,24,43)"
|
||
],
|
||
[
|
||
1.0,
|
||
"rgb(103,0,31)"
|
||
]
|
||
],
|
||
"type": "heatmap",
|
||
"x": [
|
||
"ret_5d",
|
||
"ret_10d",
|
||
"ret_21d",
|
||
"ret_42d",
|
||
"ret_63d",
|
||
"ret_126d",
|
||
"ret_189d",
|
||
"ret_252d",
|
||
"skip_recent_12_1",
|
||
"skip_recent_6_1",
|
||
"vol_21d",
|
||
"vol_63d",
|
||
"vol_126d",
|
||
"vol_252d",
|
||
"sharpe_5d",
|
||
"sharpe_10d",
|
||
"sharpe_21d",
|
||
"sharpe_42d",
|
||
"sharpe_63d",
|
||
"sharpe_126d",
|
||
"sharpe_189d",
|
||
"sharpe_252d",
|
||
"mom_accel_short",
|
||
"mom_accel_medium",
|
||
"mom_accel_long",
|
||
"vol_ratio_short",
|
||
"vol_ratio_medium",
|
||
"rsi_7",
|
||
"rsi_14",
|
||
"macd_line",
|
||
"adx_14",
|
||
"cci_14",
|
||
"cci_21",
|
||
"stoch_k",
|
||
"sma_ratio_50",
|
||
"sma_ratio_200",
|
||
"ema_ratio_26",
|
||
"bb_pctb_20",
|
||
"natr_14",
|
||
"chop_14",
|
||
"hurst_100",
|
||
"max_dd_63d",
|
||
"max_dd_126d",
|
||
"vol_ratio_21d",
|
||
"vol_ratio_63d",
|
||
"obv_zscore_63d",
|
||
"dist_52w_high",
|
||
"dist_52w_low",
|
||
"corr_spy_tlt_63d",
|
||
"ret_126d_rank",
|
||
"sharpe_126d_rank",
|
||
"yield_curve_slope",
|
||
"yield_curve_zscore",
|
||
"regime_prob_stress",
|
||
"regime_transition",
|
||
"regime_log_duration",
|
||
"ffd_spy",
|
||
"ffd_qqq",
|
||
"ffd_iwm",
|
||
"ffd_efa",
|
||
"ffd_eem",
|
||
"ffd_tlt",
|
||
"ffd_gld",
|
||
"ffd_vnq",
|
||
"ffd_hyg",
|
||
"ffd_lqd",
|
||
"garch_cond_vol"
|
||
],
|
||
"y": [
|
||
"ret_5d",
|
||
"ret_10d",
|
||
"ret_21d",
|
||
"ret_42d",
|
||
"ret_63d",
|
||
"ret_126d",
|
||
"ret_189d",
|
||
"ret_252d",
|
||
"skip_recent_12_1",
|
||
"skip_recent_6_1",
|
||
"vol_21d",
|
||
"vol_63d",
|
||
"vol_126d",
|
||
"vol_252d",
|
||
"sharpe_5d",
|
||
"sharpe_10d",
|
||
"sharpe_21d",
|
||
"sharpe_42d",
|
||
"sharpe_63d",
|
||
"sharpe_126d",
|
||
"sharpe_189d",
|
||
"sharpe_252d",
|
||
"mom_accel_short",
|
||
"mom_accel_medium",
|
||
"mom_accel_long",
|
||
"vol_ratio_short",
|
||
"vol_ratio_medium",
|
||
"rsi_7",
|
||
"rsi_14",
|
||
"macd_line",
|
||
"adx_14",
|
||
"cci_14",
|
||
"cci_21",
|
||
"stoch_k",
|
||
"sma_ratio_50",
|
||
"sma_ratio_200",
|
||
"ema_ratio_26",
|
||
"bb_pctb_20",
|
||
"natr_14",
|
||
"chop_14",
|
||
"hurst_100",
|
||
"max_dd_63d",
|
||
"max_dd_126d",
|
||
"vol_ratio_21d",
|
||
"vol_ratio_63d",
|
||
"obv_zscore_63d",
|
||
"dist_52w_high",
|
||
"dist_52w_low",
|
||
"corr_spy_tlt_63d",
|
||
"ret_126d_rank",
|
||
"sharpe_126d_rank",
|
||
"yield_curve_slope",
|
||
"yield_curve_zscore",
|
||
"regime_prob_stress",
|
||
"regime_transition",
|
||
"regime_log_duration",
|
||
"ffd_spy",
|
||
"ffd_qqq",
|
||
"ffd_iwm",
|
||
"ffd_efa",
|
||
"ffd_eem",
|
||
"ffd_tlt",
|
||
"ffd_gld",
|
||
"ffd_vnq",
|
||
"ffd_hyg",
|
||
"ffd_lqd",
|
||
"garch_cond_vol"
|
||
],
|
||
"z": {
|
||
"bdata": "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",
|
||
"dtype": "f8",
|
||
"shape": "67, 67"
|
||
},
|
||
"zmax": 1,
|
||
"zmid": 0,
|
||
"zmin": -1
|
||
}
|
||
],
|
||
"layout": {
|
||
"height": 700,
|
||
"template": {
|
||
"data": {
|
||
"bar": [
|
||
{
|
||
"error_x": {
|
||
"color": "#2a3f5f"
|
||
},
|
||
"error_y": {
|
||
"color": "#2a3f5f"
|
||
},
|
||
"marker": {
|
||
"line": {
|
||
"color": "white",
|
||
"width": 0.5
|
||
},
|
||
"pattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
}
|
||
},
|
||
"type": "bar"
|
||
}
|
||
],
|
||
"barpolar": [
|
||
{
|
||
"marker": {
|
||
"line": {
|
||
"color": "white",
|
||
"width": 0.5
|
||
},
|
||
"pattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
}
|
||
},
|
||
"type": "barpolar"
|
||
}
|
||
],
|
||
"carpet": [
|
||
{
|
||
"aaxis": {
|
||
"endlinecolor": "#2a3f5f",
|
||
"gridcolor": "#C8D4E3",
|
||
"linecolor": "#C8D4E3",
|
||
"minorgridcolor": "#C8D4E3",
|
||
"startlinecolor": "#2a3f5f"
|
||
},
|
||
"baxis": {
|
||
"endlinecolor": "#2a3f5f",
|
||
"gridcolor": "#C8D4E3",
|
||
"linecolor": "#C8D4E3",
|
||
"minorgridcolor": "#C8D4E3",
|
||
"startlinecolor": "#2a3f5f"
|
||
},
|
||
"type": "carpet"
|
||
}
|
||
],
|
||
"choropleth": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"type": "choropleth"
|
||
}
|
||
],
|
||
"contour": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "contour"
|
||
}
|
||
],
|
||
"contourcarpet": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"type": "contourcarpet"
|
||
}
|
||
],
|
||
"heatmap": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "heatmap"
|
||
}
|
||
],
|
||
"histogram": [
|
||
{
|
||
"marker": {
|
||
"pattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
}
|
||
},
|
||
"type": "histogram"
|
||
}
|
||
],
|
||
"histogram2d": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "histogram2d"
|
||
}
|
||
],
|
||
"histogram2dcontour": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "histogram2dcontour"
|
||
}
|
||
],
|
||
"mesh3d": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"type": "mesh3d"
|
||
}
|
||
],
|
||
"parcoords": [
|
||
{
|
||
"line": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "parcoords"
|
||
}
|
||
],
|
||
"pie": [
|
||
{
|
||
"automargin": true,
|
||
"type": "pie"
|
||
}
|
||
],
|
||
"scatter": [
|
||
{
|
||
"fillpattern": {
|
||
"fillmode": "overlay",
|
||
"size": 10,
|
||
"solidity": 0.2
|
||
},
|
||
"type": "scatter"
|
||
}
|
||
],
|
||
"scatter3d": [
|
||
{
|
||
"line": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatter3d"
|
||
}
|
||
],
|
||
"scattercarpet": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattercarpet"
|
||
}
|
||
],
|
||
"scattergeo": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattergeo"
|
||
}
|
||
],
|
||
"scattergl": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattergl"
|
||
}
|
||
],
|
||
"scattermap": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattermap"
|
||
}
|
||
],
|
||
"scattermapbox": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scattermapbox"
|
||
}
|
||
],
|
||
"scatterpolar": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatterpolar"
|
||
}
|
||
],
|
||
"scatterpolargl": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatterpolargl"
|
||
}
|
||
],
|
||
"scatterternary": [
|
||
{
|
||
"marker": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"type": "scatterternary"
|
||
}
|
||
],
|
||
"surface": [
|
||
{
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
},
|
||
"colorscale": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"type": "surface"
|
||
}
|
||
],
|
||
"table": [
|
||
{
|
||
"cells": {
|
||
"fill": {
|
||
"color": "#EBF0F8"
|
||
},
|
||
"line": {
|
||
"color": "white"
|
||
}
|
||
},
|
||
"header": {
|
||
"fill": {
|
||
"color": "#C8D4E3"
|
||
},
|
||
"line": {
|
||
"color": "white"
|
||
}
|
||
},
|
||
"type": "table"
|
||
}
|
||
]
|
||
},
|
||
"layout": {
|
||
"annotationdefaults": {
|
||
"arrowcolor": "#2a3f5f",
|
||
"arrowhead": 0,
|
||
"arrowwidth": 1
|
||
},
|
||
"autotypenumbers": "strict",
|
||
"coloraxis": {
|
||
"colorbar": {
|
||
"outlinewidth": 0,
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"colorscale": {
|
||
"diverging": [
|
||
[
|
||
0,
|
||
"#8e0152"
|
||
],
|
||
[
|
||
0.1,
|
||
"#c51b7d"
|
||
],
|
||
[
|
||
0.2,
|
||
"#de77ae"
|
||
],
|
||
[
|
||
0.3,
|
||
"#f1b6da"
|
||
],
|
||
[
|
||
0.4,
|
||
"#fde0ef"
|
||
],
|
||
[
|
||
0.5,
|
||
"#f7f7f7"
|
||
],
|
||
[
|
||
0.6,
|
||
"#e6f5d0"
|
||
],
|
||
[
|
||
0.7,
|
||
"#b8e186"
|
||
],
|
||
[
|
||
0.8,
|
||
"#7fbc41"
|
||
],
|
||
[
|
||
0.9,
|
||
"#4d9221"
|
||
],
|
||
[
|
||
1,
|
||
"#276419"
|
||
]
|
||
],
|
||
"sequential": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
],
|
||
"sequentialminus": [
|
||
[
|
||
0.0,
|
||
"#0d0887"
|
||
],
|
||
[
|
||
0.1111111111111111,
|
||
"#46039f"
|
||
],
|
||
[
|
||
0.2222222222222222,
|
||
"#7201a8"
|
||
],
|
||
[
|
||
0.3333333333333333,
|
||
"#9c179e"
|
||
],
|
||
[
|
||
0.4444444444444444,
|
||
"#bd3786"
|
||
],
|
||
[
|
||
0.5555555555555556,
|
||
"#d8576b"
|
||
],
|
||
[
|
||
0.6666666666666666,
|
||
"#ed7953"
|
||
],
|
||
[
|
||
0.7777777777777778,
|
||
"#fb9f3a"
|
||
],
|
||
[
|
||
0.8888888888888888,
|
||
"#fdca26"
|
||
],
|
||
[
|
||
1.0,
|
||
"#f0f921"
|
||
]
|
||
]
|
||
},
|
||
"colorway": [
|
||
"#636efa",
|
||
"#EF553B",
|
||
"#00cc96",
|
||
"#ab63fa",
|
||
"#FFA15A",
|
||
"#19d3f3",
|
||
"#FF6692",
|
||
"#B6E880",
|
||
"#FF97FF",
|
||
"#FECB52"
|
||
],
|
||
"font": {
|
||
"color": "#2a3f5f"
|
||
},
|
||
"geo": {
|
||
"bgcolor": "white",
|
||
"lakecolor": "white",
|
||
"landcolor": "white",
|
||
"showlakes": true,
|
||
"showland": true,
|
||
"subunitcolor": "#C8D4E3"
|
||
},
|
||
"hoverlabel": {
|
||
"align": "left"
|
||
},
|
||
"hovermode": "closest",
|
||
"mapbox": {
|
||
"style": "light"
|
||
},
|
||
"paper_bgcolor": "white",
|
||
"plot_bgcolor": "white",
|
||
"polar": {
|
||
"angularaxis": {
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": ""
|
||
},
|
||
"bgcolor": "white",
|
||
"radialaxis": {
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"scene": {
|
||
"xaxis": {
|
||
"backgroundcolor": "white",
|
||
"gridcolor": "#DFE8F3",
|
||
"gridwidth": 2,
|
||
"linecolor": "#EBF0F8",
|
||
"showbackground": true,
|
||
"ticks": "",
|
||
"zerolinecolor": "#EBF0F8"
|
||
},
|
||
"yaxis": {
|
||
"backgroundcolor": "white",
|
||
"gridcolor": "#DFE8F3",
|
||
"gridwidth": 2,
|
||
"linecolor": "#EBF0F8",
|
||
"showbackground": true,
|
||
"ticks": "",
|
||
"zerolinecolor": "#EBF0F8"
|
||
},
|
||
"zaxis": {
|
||
"backgroundcolor": "white",
|
||
"gridcolor": "#DFE8F3",
|
||
"gridwidth": 2,
|
||
"linecolor": "#EBF0F8",
|
||
"showbackground": true,
|
||
"ticks": "",
|
||
"zerolinecolor": "#EBF0F8"
|
||
}
|
||
},
|
||
"shapedefaults": {
|
||
"line": {
|
||
"color": "#2a3f5f"
|
||
}
|
||
},
|
||
"ternary": {
|
||
"aaxis": {
|
||
"gridcolor": "#DFE8F3",
|
||
"linecolor": "#A2B1C6",
|
||
"ticks": ""
|
||
},
|
||
"baxis": {
|
||
"gridcolor": "#DFE8F3",
|
||
"linecolor": "#A2B1C6",
|
||
"ticks": ""
|
||
},
|
||
"bgcolor": "white",
|
||
"caxis": {
|
||
"gridcolor": "#DFE8F3",
|
||
"linecolor": "#A2B1C6",
|
||
"ticks": ""
|
||
}
|
||
},
|
||
"title": {
|
||
"x": 0.05
|
||
},
|
||
"xaxis": {
|
||
"automargin": true,
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": "",
|
||
"title": {
|
||
"standoff": 15
|
||
},
|
||
"zerolinecolor": "#EBF0F8",
|
||
"zerolinewidth": 2
|
||
},
|
||
"yaxis": {
|
||
"automargin": true,
|
||
"gridcolor": "#EBF0F8",
|
||
"linecolor": "#EBF0F8",
|
||
"ticks": "",
|
||
"title": {
|
||
"standoff": 15
|
||
},
|
||
"zerolinecolor": "#EBF0F8",
|
||
"zerolinewidth": 2
|
||
}
|
||
}
|
||
},
|
||
"title": {
|
||
"text": "Feature Correlation Matrix (157 pairs above 0.7)"
|
||
},
|
||
"width": 800
|
||
}
|
||
},
|
||
"image/png": "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"
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"fig = go.Figure(\n",
|
||
" data=go.Heatmap(\n",
|
||
" z=corr_matrix.values,\n",
|
||
" x=corr_matrix.columns.tolist(),\n",
|
||
" y=corr_matrix.columns.tolist(),\n",
|
||
" colorscale=\"RdBu_r\",\n",
|
||
" zmid=0,\n",
|
||
" zmin=-1,\n",
|
||
" zmax=1,\n",
|
||
" )\n",
|
||
")\n",
|
||
"fig.update_layout(\n",
|
||
" title=f\"Feature Correlation Matrix ({len(high_corr_pairs)} pairs above 0.7)\",\n",
|
||
" template=\"plotly_white\",\n",
|
||
" height=700,\n",
|
||
" width=800,\n",
|
||
")\n",
|
||
"fig.show()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "352fea32",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"**Interpretation**: The correlation heatmap reveals clusters of highly\n",
|
||
"correlated features — particularly within the momentum (multi-horizon\n",
|
||
"returns) and volatility (multi-horizon realized vol) families. Downstream\n",
|
||
"modeling in Ch11 should use clustering or PCA within families to reduce\n",
|
||
"redundancy, or rely on tree-based models that handle correlated inputs\n",
|
||
"natively."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "cf7ed89c",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"## 6. Triage & Handoff\n",
|
||
"\n",
|
||
"Each feature receives a triage decision:\n",
|
||
"\n",
|
||
"| Decision | Criteria |\n",
|
||
"|----------|----------|\n",
|
||
"| **PROCEED** | FDR-significant at 5%, OR sign consistent across > 60% of folds AND abs(IC) > 0.01 |\n",
|
||
"| **STOP** | Correctness failure (coverage < 70% or staleness > 50%) |\n",
|
||
"| **REVISE** | Everything else — evaluate in multivariate context in Ch11 |\n",
|
||
"\n",
|
||
"Date-level features are triaged as REVISE with a note, since their value\n",
|
||
"lies in regime-conditional interactions rather than standalone cross-sectional IC."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 22,
|
||
"id": "29a99549",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:08.200598Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:08.200363Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:08.206154Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:08.204932Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"triage = {}\n",
|
||
"for feat in all_feature_cols:\n",
|
||
" if not correctness[feat]:\n",
|
||
" triage[feat] = (\"STOP\", \"correctness_fail\")\n",
|
||
" continue\n",
|
||
"\n",
|
||
" if feat in date_level_features:\n",
|
||
" triage[feat] = (\"REVISE\", \"date_level_feature\")\n",
|
||
" continue\n",
|
||
"\n",
|
||
" if feat not in ic_results:\n",
|
||
" triage[feat] = (\"REVISE\", \"insufficient_data\")\n",
|
||
" continue\n",
|
||
"\n",
|
||
" is_fdr_sig = feat in fdr_sig_set\n",
|
||
" sign_con = fold_stats.get(feat, {}).get(\"sign_consistency\", 0)\n",
|
||
" abs_ic = abs(ic_results[feat][\"mean_ic\"])\n",
|
||
"\n",
|
||
" if is_fdr_sig:\n",
|
||
" triage[feat] = (\"PROCEED\", \"fdr_significant\")\n",
|
||
" elif sign_con >= 0.60 and abs_ic >= IC_THRESHOLD:\n",
|
||
" triage[feat] = (\"PROCEED\", \"stable_and_above_threshold\")\n",
|
||
" else:\n",
|
||
" triage[feat] = (\"REVISE\", \"not_significant_standalone\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 23,
|
||
"id": "ea0a1b9f",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:08.239947Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:08.239355Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:08.289110Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:08.284864Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Triage ledger saved: case_studies/etfs/evaluation/triage_ledger.parquet\n",
|
||
"shape: (3, 2)\n",
|
||
"┌──────────┬─────┐\n",
|
||
"│ decision ┆ len │\n",
|
||
"│ --- ┆ --- │\n",
|
||
"│ str ┆ u32 │\n",
|
||
"╞══════════╪═════╡\n",
|
||
"│ PROCEED ┆ 17 │\n",
|
||
"│ REVISE ┆ 50 │\n",
|
||
"│ STOP ┆ 4 │\n",
|
||
"└──────────┴─────┘\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Build triage ledger\n",
|
||
"ledger_rows = []\n",
|
||
"for feat in all_feature_cols:\n",
|
||
" decision, note = triage[feat]\n",
|
||
" row = {\n",
|
||
" \"feature\": feat,\n",
|
||
" \"family\": families.get(feat, \"other\"),\n",
|
||
" \"source\": \"temporal\" if feat in temporal_cols else \"financial\",\n",
|
||
" \"ic_mean\": ic_results.get(feat, {}).get(\"mean_ic\"),\n",
|
||
" \"hac_t\": ic_results.get(feat, {}).get(\"t_stat\"),\n",
|
||
" \"hac_p\": ic_results.get(feat, {}).get(\"p_value\"),\n",
|
||
" \"fdr_p\": None,\n",
|
||
" \"fdr_sig\": False,\n",
|
||
" \"sign_consistency\": fold_stats.get(feat, {}).get(\"sign_consistency\"),\n",
|
||
" \"worst_fold_ic\": fold_stats.get(feat, {}).get(\"worst_fold_ic\"),\n",
|
||
" \"monotonicity\": monotonicity_scores.get(feat),\n",
|
||
" \"coverage\": coverage[feat],\n",
|
||
" \"staleness\": staleness[feat],\n",
|
||
" \"decision\": decision,\n",
|
||
" \"note\": note,\n",
|
||
" }\n",
|
||
" match = eval_summary.filter(pl.col(\"feature\") == feat)\n",
|
||
" if len(match) > 0:\n",
|
||
" row[\"fdr_p\"] = float(match[\"fdr_p\"][0])\n",
|
||
" row[\"fdr_sig\"] = bool(match[\"fdr_sig\"][0])\n",
|
||
" ledger_rows.append(row)\n",
|
||
"\n",
|
||
"triage_ledger = pl.DataFrame(ledger_rows)\n",
|
||
"triage_ledger.write_parquet(EVAL_DIR / \"triage_ledger.parquet\")\n",
|
||
"print(f\"Triage ledger saved: {EVAL_DIR / 'triage_ledger.parquet'}\")\n",
|
||
"print(triage_ledger.group_by(\"decision\").len().sort(\"decision\"))"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 24,
|
||
"id": "e45f68e7",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:08.316048Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:08.315782Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:08.435668Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:08.428494Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"IC time series saved: case_studies/etfs/evaluation/ic_timeseries.parquet\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"# Save IC time series (long format)\n",
|
||
"ic_ts_frames = []\n",
|
||
"for feat, ts in ic_timeseries.items():\n",
|
||
" ic_ts_frames.append(ts.with_columns(pl.lit(feat).alias(\"feature\")))\n",
|
||
"\n",
|
||
"if ic_ts_frames:\n",
|
||
" ic_ts_all = pl.concat(ic_ts_frames)\n",
|
||
" ic_ts_all.write_parquet(EVAL_DIR / \"ic_timeseries.parquet\")\n",
|
||
" print(f\"IC time series saved: {EVAL_DIR / 'ic_timeseries.parquet'}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 25,
|
||
"id": "97600726",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:08.466389Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:08.465933Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:08.470521Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:08.469734Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"# Write results JSON\n",
|
||
"proceed_features = sorted(f for f, (d, _) in triage.items() if d == \"PROCEED\")\n",
|
||
"revise_features = [f for f, (d, _) in triage.items() if d == \"REVISE\"]\n",
|
||
"stop_features = [f for f, (d, _) in triage.items() if d == \"STOP\"]\n",
|
||
"\n",
|
||
"sorted_by_ic = sorted(ic_results.items(), key=lambda x: x[1].get(\"mean_ic\") or 0, reverse=True)\n",
|
||
"best = sorted_by_ic[0] if sorted_by_ic else (\"n/a\", {})\n",
|
||
"worst = sorted_by_ic[-1] if sorted_by_ic else (\"n/a\", {})"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 26,
|
||
"id": "006b4efd",
|
||
"metadata": {
|
||
"execution": {
|
||
"iopub.execute_input": "2026-05-15T05:25:08.499614Z",
|
||
"iopub.status.busy": "2026-05-15T05:25:08.499144Z",
|
||
"iopub.status.idle": "2026-05-15T05:25:08.507371Z",
|
||
"shell.execute_reply": "2026-05-15T05:25:08.506064Z"
|
||
},
|
||
"tags": []
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"\n",
|
||
"============================================================\n",
|
||
"TRIAGE SUMMARY: etfs\n",
|
||
"============================================================\n",
|
||
" PROCEED: 17 features\n",
|
||
" REVISE: 50 features\n",
|
||
" STOP: 4 features\n",
|
||
"\n",
|
||
"PROMOTED (PROCEED) features:\n",
|
||
" dist_52w_low IC=+0.0764 t=4.64 [distance]\n",
|
||
" garch_cond_vol IC=+0.0496 t=2.54 [temporal_other]\n",
|
||
" natr_14 IC=+0.0572 t=2.92 [volatility]\n",
|
||
" ret_126d IC=+0.0234 t=1.35 [momentum]\n",
|
||
" ret_126d_rank IC=+0.0234 t=1.35 [momentum]\n",
|
||
" ret_189d IC=+0.0411 t=2.29 [momentum]\n",
|
||
" ret_252d IC=+0.0403 t=2.22 [momentum]\n",
|
||
" ret_63d IC=+0.0201 t=1.19 [momentum]\n",
|
||
" sharpe_189d IC=+0.0160 t=0.96 [risk_adj_momentum]\n",
|
||
" skip_recent_12_1 IC=+0.0418 t=2.29 [momentum]\n",
|
||
" skip_recent_6_1 IC=+0.0241 t=1.37 [momentum]\n",
|
||
" sma_ratio_200 IC=+0.0335 t=1.95 [trend]\n",
|
||
" vol_126d IC=+0.0564 t=2.78 [volatility]\n",
|
||
" vol_21d IC=+0.0507 t=2.62 [volatility]\n",
|
||
" vol_252d IC=+0.0589 t=2.87 [volatility]\n",
|
||
" vol_63d IC=+0.0509 t=2.55 [volatility]\n",
|
||
" vol_ratio_63d IC=+0.0105 t=2.12 [vol_ratio]\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"print(f\"\\n{'=' * 60}\")\n",
|
||
"print(f\"TRIAGE SUMMARY: {CASE_STUDY_ID}\")\n",
|
||
"print(f\"{'=' * 60}\")\n",
|
||
"print(f\" PROCEED: {len(proceed_features)} features\")\n",
|
||
"print(f\" REVISE: {len(revise_features)} features\")\n",
|
||
"print(f\" STOP: {len(stop_features)} features\")\n",
|
||
"print(\"\\nPROMOTED (PROCEED) features:\")\n",
|
||
"for f in proceed_features:\n",
|
||
" ic = ic_results[f][\"mean_ic\"]\n",
|
||
" t = ic_results[f][\"t_stat\"]\n",
|
||
" print(f\" {f:40s} IC={ic:+.4f} t={t:.2f} [{families.get(f, '?')}]\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "c19b103c",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"### Quality Gate Verdict\n",
|
||
"\n",
|
||
"**Fit for modeling.** The triage promotes a broad set of features to PROCEED,\n",
|
||
"concentrated in the momentum and risk-adjusted momentum families. IC magnitudes\n",
|
||
"are moderate (0.03--0.04 range), consistent with monthly-horizon cross-sectional\n",
|
||
"equity signals over a 100-ETF universe. Date-level regime and macro features\n",
|
||
"are excluded from cross-sectional IC but carry forward as conditioning variables\n",
|
||
"for tree-based and interaction models in Ch11+."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "b0fb9b73",
|
||
"metadata": {
|
||
"tags": []
|
||
},
|
||
"source": [
|
||
"## Key Takeaways\n",
|
||
"\n",
|
||
"1. **FDR correction is essential**: With 50+ features, naive p-values\n",
|
||
" dramatically overstate the number of true signals. The inflation factor\n",
|
||
" quantifies the gap between naive and corrected significance.\n",
|
||
"\n",
|
||
"2. **HAC adjustment matters**: Overlapping 21-day returns create\n",
|
||
" autocorrelation in the IC time series. Newey-West standard errors\n",
|
||
" with bandwidth = 21 properly account for this.\n",
|
||
"\n",
|
||
"3. **Date-level features require special handling**: Regime and macro\n",
|
||
" features have zero cross-sectional IC by construction. Their value\n",
|
||
" emerges through conditional interactions in Ch11 modeling.\n",
|
||
"\n",
|
||
"4. **Redundancy within families**: Momentum and volatility features are\n",
|
||
" highly correlated. Downstream modeling should cluster or select\n",
|
||
" representative features from each family.\n",
|
||
"\n",
|
||
"**Next**: `05_linear.py` (Ch11) uses the triage ledger and promoted\n",
|
||
"features for ridge/lasso baseline modeling."
|
||
]
|
||
}
|
||
],
|
||
"metadata": {
|
||
"kernelspec": {
|
||
"display_name": "Python 3",
|
||
"language": "python",
|
||
"name": "python3"
|
||
},
|
||
"language_info": {
|
||
"codemirror_mode": {
|
||
"name": "ipython",
|
||
"version": 3
|
||
},
|
||
"file_extension": ".py",
|
||
"mimetype": "text/x-python",
|
||
"name": "python",
|
||
"nbconvert_exporter": "python",
|
||
"pygments_lexer": "ipython3",
|
||
"version": "3.14.3"
|
||
},
|
||
"papermill": {
|
||
"default_parameters": {},
|
||
"duration": 236.669733,
|
||
"end_time": "2026-05-15T05:25:11.493581+00:00",
|
||
"environment_variables": {},
|
||
"exception": null,
|
||
"input_path": "case_studies/etfs/05_evaluation.ipynb",
|
||
"output_path": "case_studies/etfs/05_evaluation.ipynb",
|
||
"parameters": {},
|
||
"start_time": "2026-05-15T05:21:14.823848+00:00",
|
||
"version": "2.7.0"
|
||
}
|
||
},
|
||
"nbformat": 4,
|
||
"nbformat_minor": 5
|
||
}
|