38.4 GB
33,742 files
Updated 16 days ago
Name
Size
downloads
earning_transcripts
snapshots
stock_list
stock_price_daily
stock_price_daily_adjusted_volume
stock_price_daily_with_marketcap
README.md2.09 kB
xet
blacklist_data_error.csv1.02 kB
xet
blacklist_data_error_README.md3.08 kB
xet
half_days.csv489 Bytes
xet
price_jumps.csv122 kB
xet
price_jumps_validated.csv21.1 kB
xet
README.md

data/

Data directory for two-alpha-trading. All price data sourced from FMP, universe from Tiger Brokers US equity list.

Directory Structure

data/
├── stock_price_daily/          # 2550 symbols, raw daily OHLCV parquets
├── stock_price_daily_adjusted_volume/  # Same with half-day volume adjustment
├── snapshots/
│   ├── normal/                 # Per-day snapshots (OHLCV + 171 Alpha191)
│   ├── adjusted_volume/        # Same, from adjusted-volume prices
│   ├── normalized/             # v1: spot close/volume normalizers
│   ├── normalized_adjvol/      # v1: same, from adjusted-volume prices
│   └── normalized_v2/          # v2: rolling-mean normalizers (recommended)
├── blacklist_data_error.csv    # 39 excluded symbols
├── blacklist_data_error_README.md
├── half_days.csv               # 44 half-trading day dates
├── price_jumps.csv             # 1147 detected price jumps (320 symbols)
└── price_jumps_validated.csv   # Polygon.io cross-validation results

Key Files

File Description
blacklist_data_error.csv 39 symbols excluded: 30 for price jump data errors, 9 for alpha overflow
half_days.csv 44 US market half-trading days (early close)
price_jumps.csv All detected |daily return| > 50% events, classified as likely_split / likely_event / data_error
price_jumps_validated.csv Cross-validation of FMP prices against Polygon.io

Data Pipeline

FMP API → stock_price_daily/ (raw)
             ├── adjust_half_day_volume.py → stock_price_daily_adjusted_volume/
             ├── detect_price_jumps.py → price_jumps.csv → blacklist
             └── generate_snapshots.py → snapshots/normal/, adjusted_volume/
                    ├── normalize_snapshots.py → snapshots/normalized/, normalized_adjvol/
                    └── normalize_snapshots_v2.py → snapshots/normalized_v2/

See each subdirectory's README.md for detailed schema and known issues.

Total size
38.4 GB
Files
33,742
Last updated
Jul 24
Pre-warmed CDN
US EU US EU

Contributors