Duplicate from Torch-Trade/btcusdt_perp_bookticker_features_1m_05_2023_to_03_2024

.gitattributes

@@ -0,0 +1,60 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.avro filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.lz4 filter=lfs diff=lfs merge=lfs -text
+*.mds filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+# Audio files - uncompressed
+*.pcm filter=lfs diff=lfs merge=lfs -text
+*.sam filter=lfs diff=lfs merge=lfs -text
+*.raw filter=lfs diff=lfs merge=lfs -text
+# Audio files - compressed
+*.aac filter=lfs diff=lfs merge=lfs -text
+*.flac filter=lfs diff=lfs merge=lfs -text
+*.mp3 filter=lfs diff=lfs merge=lfs -text
+*.ogg filter=lfs diff=lfs merge=lfs -text
+*.wav filter=lfs diff=lfs merge=lfs -text
+# Image files - uncompressed
+*.bmp filter=lfs diff=lfs merge=lfs -text
+*.gif filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text
+*.tiff filter=lfs diff=lfs merge=lfs -text
+# Image files - compressed
+*.jpg filter=lfs diff=lfs merge=lfs -text
+*.jpeg filter=lfs diff=lfs merge=lfs -text
+*.webp filter=lfs diff=lfs merge=lfs -text
+# Video files - compressed
+*.mp4 filter=lfs diff=lfs merge=lfs -text
+*.webm filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,118 @@
+---
+license: mit
+task_categories:
+- time-series-forecasting
+tags:
+- finance
+- crypto
+- bitcoin
+- btcusdt
+- binance
+- futures
+- microstructure
+- order-book
+- bookticker
+pretty_name: BTCUSDT Perpetual Futures L1 BookTicker Features (1m)
+size_categories:
+- 100K<n<1M
+---
+
+# BTCUSDT Perpetual Futures L1 BookTicker Features (1m)
+
+1-minute aggregated features derived from Binance UM Futures **bookTicker** tick data
+(best bid/ask price + quantity updates). Captures spread, top-of-book imbalance,
+microprice, and quote update intensity — the level-1 microstructure layer.
+
+## Coverage
+
+- **Range**: 2023-05-16 11:49 → 2024-03-31 23:59 (UTC)
+- **Cadence**: 1 minute
+- **Rows**: 460,265
+- **Completeness vs 1-min grid**: 99.726%
+
+### ⚠️ Temporal gap — Binance stopped publishing bookTicker historical dumps
+
+Binance publishes bookTicker as monthly files at
+`https://data.binance.vision/data/futures/um/monthly/bookTicker/BTCUSDT/`. Available files
+cover **2023-05 through 2024-03 only** (11 months). The 2024-04 file is published but
+is 0 bytes (corrupt). All files 2024-05 onwards return HTTP 404. No daily bookTicker
+files are published.
+
+**If you need bookTicker features for 2024-04 onwards, you must collect them live
+yourself from the Binance WebSocket `@bookTicker` stream**. This dataset cannot cover
+that period.
+
+## Columns
+
+| Column | Description |
+|---|---|
+| `timestamp` | 1-min bar open time (UTC) |
+| `bt_spread_bps_close` | Spread at last tick of bar, in basis points: `(ask − bid)/mid × 10000` |
+| `bt_spread_bps_twap` | Mean spread across all ticks within the bar |
+| `bt_bid_qty_close` | Best bid quantity at last tick |
+| `bt_ask_qty_close` | Best ask quantity at last tick |
+| `bt_imbalance_close` | `(bid_qty − ask_qty) / (bid_qty + ask_qty)` at last tick. Range [-1, 1] |
+| `bt_imbalance_twap` | Mean imbalance across all ticks within the bar |
+| `bt_microprice_close` | `(ask × bid_qty + bid × ask_qty) / (bid_qty + ask_qty)` at last tick |
+| `bt_microprice_premium_close` | `(microprice − mid) / mid` — signed pressure proxy |
+| `bt_update_rate` | Count of bookTicker events within the bar (quote activity intensity) |
+
+### Feature meaning
+
+- **Imbalance** > 0: more size on the bid; canonical short-term predictor of upward price drift
+  (Cartea/Jaimungal/Penalva, *Algorithmic and High-Frequency Trading*, Cambridge 2015).
+- **TWAP vs close**: TWAP averages over the whole bar; close is the latest snapshot.
+  Large divergence between TWAP and close indicates persistent vs transient imbalance.
+- **Microprice premium**: the microprice sits between bid and ask weighted by sizes.
+  When positive, pressure favors upward moves; when negative, downward. Signed.
+- **Update rate**: more quote churn correlates with information arrival / elevated
+  volatility.
+
+## Source
+
+Binance publishes per-tick bookTicker events as monthly zips. This dataset is built
+by streaming each monthly file, computing per-minute aggregations, discarding the raw,
+and concatenating. Build script:
+`scripts/build_datasets/build_bookticker.py` in the `torchtrade-experiments` repo.
+
+## Usage
+
+```python
+from datasets import load_dataset
+import pandas as pd
+
+ds = load_dataset("Torch-Trade/btcusdt_perp_bookticker_features_1m_05_2023_to_03_2024", split="train")
+bt = ds.to_pandas().sort_values("timestamp").reset_index(drop=True)
+print(bt.head())
+```
+
+### Joining with 1-minute OHLCV
+
+```python
+from datasets import load_dataset
+import pandas as pd
+
+ohlcv = load_dataset(
+    "Torch-Trade/btcusdt_perp_1m_05_2021_to_02_2026", split="train"
+).to_pandas()
+bt = load_dataset("Torch-Trade/btcusdt_perp_bookticker_features_1m_05_2023_to_03_2024", split="train").to_pandas()
+
+ohlcv["timestamp"] = pd.to_datetime(ohlcv["timestamp"])
+bt["timestamp"] = pd.to_datetime(bt["timestamp"])
+
+# Both are 1-min native — direct inner join on timestamp
+joined = ohlcv.merge(bt, on="timestamp", how="inner")
+```
+
+Expected `joined` row count ≈ 482,000 across the 11-month overlap window.
+
+### Derived features at training time
+
+- `imb_momentum = bt_imbalance_close.diff(5)` — 5-min change in imbalance
+- `spread_zscore = (bt_spread_bps_close - rolling_mean_60) / rolling_std_60`
+- `micro_prem_sign_persist` — fraction of last N bars where microprice_premium had the same sign
+
+## License
+
+Data © Binance. Redistributed under fair use for research purposes.
+See https://data.binance.vision/ for upstream licensing.

data/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:274eb8e87c7d7185a0162271144b30a0e387ae496fe657c6af83833448f08624
+size 28423067