Duplicate from SII-WANGZJ/Polymarket_data

Co-authored-by: SII-WANGZJ <SII-WANGZJ@users.noreply.huggingface.co>

.gitattributes

@@ -0,0 +1,59 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow 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

DATA_DESCRIPTION.md

@@ -0,0 +1,449 @@
+# Polymarket Dataset Description
+
+Complete guide to the 5 parquet files in this dataset.
+
+---
+
+## 📁 File Overview
+
+| File | Size | Description | Use Case |
+|------|------|-------------|----------|
+| `orderfilled.parquet` | 31GB | Raw blockchain events | Complete historical record |
+| `trades.parquet` | 32GB | Processed trading data | Market analysis, price tracking |
+| `markets.parquet` | 68MB | Market metadata | Market info, token mapping |
+| `quant.parquet` | 21GB | Quantitative trading data | Trading algorithms, backtesting |
+| `users.parquet` | 23GB | User behavior data | User analysis, wallet tracking |
+
+---
+
+## 1️⃣ orderfilled.parquet (31GB)
+
+**Raw OrderFilled events from Polygon blockchain**
+
+### Description
+Original blockchain events decoded from two Polymarket exchange contracts:
+- CTF Exchange: `0x4bfb41d5b3570defd03c39a9a4d8de6bd8b8982e`
+- NegRisk CTF Exchange: `0xc5d563a36ae78145c45a50134d48a1215220f80a`
+
+### Schema
+```
+timestamp           int64      # Unix timestamp
+datetime            string     # Human-readable datetime
+block_number        int64      # Blockchain block number
+transaction_hash    string     # Transaction hash
+log_index           int64      # Log index within transaction
+contract            string     # Contract name
+maker               string     # Maker address
+taker               string     # Taker address
+maker_asset_id      string     # Maker's asset ID (uint256 as string)
+taker_asset_id      string     # Taker's asset ID (uint256 as string)
+maker_amount_filled int64      # Maker's filled amount (wei)
+taker_amount_filled int64      # Taker's filled amount (wei)
+maker_fee           int64      # Maker fee (wei)
+taker_fee           int64      # Taker fee (wei)
+protocol_fee        int64      # Protocol fee (wei)
+order_hash          string     # Order hash
+```
+
+### Key Features
+- ✅ Complete blockchain data with no missing fields
+- ✅ Includes block_number for time-series analysis
+- ✅ Includes all fees (maker, taker, protocol)
+- ✅ Contains order_hash for order tracking
+
+### Use Cases
+- Full historical blockchain analysis
+- Fee analysis
+- Order matching studies
+- Raw event processing
+
+---
+
+## 2️⃣ trades.parquet (32GB)
+
+**Processed trading data with market linkage**
+
+### Description
+Enhanced version of orderfilled events with:
+- Market ID linkage (from token to market)
+- Trade direction analysis (BUY/SELL)
+- Price calculation
+- USD amount extraction
+
+### Schema
+All fields from `orderfilled.parquet`, plus:
+```
+market_id           string     # Linked market ID
+condition_id        string     # Condition ID
+token_id            string     # Non-USDC token ID
+answer              string     # Market outcome (YES/NO/etc.)
+nonusdc_side        string     # Which side has token (token1/token2)
+maker_direction     string     # Maker's direction (BUY/SELL)
+taker_direction     string     # Taker's direction (BUY/SELL)
+price               float64    # Trade price (0-1)
+token_amount        int64      # Token amount (in wei)
+usd_amount          int64      # USDC amount (in wei)
+```
+
+### Key Features
+- ✅ Auto-linked to market metadata
+- ✅ Trade direction calculated for both sides
+- ✅ Price normalized to 0-1 range
+- ✅ Preserves all original blockchain fields
+
+### Use Cases
+- Market price analysis
+- Trading volume by market
+- Direction-based analysis
+- General trading analytics
+
+---
+
+## 3️⃣ markets.parquet (68MB)
+
+**Market metadata from Gamma API**
+
+### Description
+Complete market information including:
+- Market questions
+- Outcome names
+- Token IDs
+- Resolution status
+
+### Schema
+```
+market_id           string     # Unique market ID
+question            string     # Market question
+description         string     # Market description
+outcomes            list       # List of outcome names
+tokens              list       # List of token IDs
+volume              float64    # Total trading volume
+liquidity           float64    # Market liquidity
+resolved            bool       # Is market resolved?
+resolution          string     # Resolution outcome
+created_at          timestamp  # Market creation time
+end_date            timestamp  # Market end date
+```
+
+### Use Cases
+- Market information lookup
+- Token ID to market mapping
+- Market category analysis
+
+---
+
+## 4️⃣ quant.parquet (21GB) - Quantitative Trading Data
+
+**Cleaned trades with unified token perspective and direction**
+
+### Description
+Optimized version of `trades.parquet` for quantitative analysis with:
+1. **Unified Token Perspective**: All trades normalized to YES token (token1)
+2. **Unified Direction**: All trades converted to BUY perspective
+3. **Contract Filtering**: Removed contract-to-contract trades
+
+### Data Cleaning Process
+
+#### Step 1: Filter Invalid Data
+```python
+# Remove trades with NaN prices
+df = df[~df['price'].isna()]
+
+# Remove contract addresses as taker (internal contract operations)
+contract_addresses = {
+    '0x4bfb41d5b3570defd03c39a9a4d8de6bd8b8982e',  # CTF Exchange
+    '0xc5d563a36ae78145c45a50134d48a1215220f80a'   # NegRisk CTF Exchange
+}
+df = df[~df['taker'].str.lower().isin(contract_addresses)]
+```
+
+#### Step 2: Unified Token Perspective (YES/token1)
+```python
+# If trade involves token2 (NO token), flip the price
+is_token2 = df['nonusdc_side'] == 'token2'
+df.loc[is_token2, 'price'] = 1 - df.loc[is_token2, 'price']
+df['nonusdc_side'] = 'token1'  # All normalized to token1/YES
+
+# Example:
+# Original: token2, price=0.3 → Normalized: token1, price=0.7
+```
+
+**Why?** In prediction markets:
+- token1 = YES outcome (probability)
+- token2 = NO outcome (1 - probability)
+
+By normalizing all trades to YES perspective:
+- price = 0.7 means 70% probability of YES
+- Easier to compare across all trades
+- Consistent time-series analysis
+
+#### Step 3: Keep All Original Fields
+Unlike `users.parquet`, `quant.parquet` preserves:
+- Both maker and taker addresses
+- Both maker_direction and taker_direction
+- All blockchain metadata
+
+### Schema
+Same as `trades.parquet`, but with:
+```
+nonusdc_side        string     # Always 'token1' (unified)
+price               float64    # YES token price (unified)
+```
+
+### Key Features
+- ✅ Unified YES token perspective for all trades
+- ✅ Contract trades filtered out
+- ✅ Maintains maker/taker structure
+- ✅ Preserves all original fields
+- ✅ Clean data for backtesting
+
+### Use Cases
+- **Algorithmic Trading**: Consistent price signals
+- **Backtesting**: Unified direction simplifies strategy testing
+- **Price Prediction**: Model YES probability trends
+- **Market Making**: Analyze bid-ask spreads
+- **Volume Analysis**: Track real trading activity (no contracts)
+
+---
+
+## 5️⃣ users.parquet (23GB) - User Behavior Data
+
+**User-centric view with split maker/taker records**
+
+### Description
+Transformed view of trades optimized for user behavior analysis:
+1. **Split Records**: Each trade becomes 2 records (maker + taker)
+2. **Unified Token Perspective**: All normalized to YES token
+3. **Unified Direction**: All trades converted to BUY, negative amounts for sells
+4. **Sorted by User**: Easy to analyze individual user trajectories
+
+### Data Cleaning Process
+
+#### Step 1: Filter Invalid Data (Same as quant)
+```python
+# Remove NaN prices and contract trades
+df = df[~df['price'].isna()]
+df = df[~df['taker'].str.lower().isin(contract_addresses)]
+```
+
+#### Step 2: Split Maker and Taker into Separate Records
+```python
+# Common fields for both sides
+common = ['timestamp', 'datetime', 'block_number', 'transaction_hash',
+          'market_id', 'price', 'token_amount', 'usd_amount']
+
+# Maker record
+maker_df = df[common + ['maker', 'maker_direction']].copy()
+maker_df = maker_df.rename(columns={'maker': 'user', 'maker_direction': 'direction'})
+maker_df['role'] = 'maker'
+
+# Taker record
+taker_df = df[common + ['taker', 'taker_direction']].copy()
+taker_df = taker_df.rename(columns={'taker': 'user', 'taker_direction': 'direction'})
+taker_df['role'] = 'taker'
+
+# Concatenate
+result = pd.concat([maker_df, taker_df])
+
+# Original 1 trade → 2 records (expansion ratio ~2x)
+```
+
+#### Step 3: Unified Token Perspective (YES/token1)
+```python
+# If token2 trade, flip price (same as quant)
+is_token2 = result['nonusdc_side'] == 'token2'
+result.loc[is_token2, 'price'] = 1 - result.loc[is_token2, 'price']
+```
+
+#### Step 4: Unified Direction to BUY
+```python
+# Convert all SELL to BUY with negative token_amount
+is_sell = result['direction'] == 'SELL'
+result.loc[is_sell, 'token_amount'] = -result.loc[is_sell, 'token_amount']
+result['direction'] = 'BUY'  # All records are now 'BUY'
+
+# Example:
+# Original SELL: direction='SELL', token_amount=100
+# Unified BUY:  direction='BUY',  token_amount=-100
+```
+
+**Why?**
+- **BUY** (token_amount > 0): User bought YES tokens, spent USDC
+- **SELL** (token_amount < 0): User sold YES tokens, received USDC
+- Single direction simplifies aggregation: `sum(token_amount)` = net position
+
+#### Step 5: Sort by User and Time
+```python
+result = result.sort_values(['user', 'timestamp'])
+```
+
+### Schema
+```
+timestamp           int64      # Unix timestamp
+datetime            string     # Human-readable datetime
+block_number        int64      # Block number
+transaction_hash    string     # Transaction hash
+event_id            string     # Unique event identifier
+market_id           string     # Market ID
+condition_id        string     # Condition ID
+user                string     # User address (was maker/taker)
+role                string     # 'maker' or 'taker'
+price               float64    # YES token price (unified)
+token_amount        int64      # Token amount (negative if originally SELL)
+usd_amount          int64      # USDC amount (wei)
+```
+
+### Key Features
+- ✅ Each user appears in their own records
+- ✅ Easy to track user trading history
+- ✅ Unified BUY direction with signed amounts
+- ✅ Sorted by user → timestamp for sequential analysis
+- ✅ 2x expansion ratio (1 trade → 2 records)
+
+### Use Cases
+- **User Profiling**: Track individual user strategies
+- **Wallet Analysis**: PnL calculation per user
+- **Cohort Analysis**: User behavior segmentation
+- **Position Tracking**: Sum token_amount = net position
+- **Trading Pattern Detection**: Identify market makers, arbitrageurs
+- **User Journey**: Sequential trade analysis
+
+### Example: Calculate User Net Position
+```python
+import pandas as pd
+
+users = pd.read_parquet('users.parquet')
+
+# Net position per user per market
+position = users.groupby(['user', 'market_id'])['token_amount'].sum()
+
+# User with long position (bought more than sold)
+positive_position = position[position > 0]
+
+# User with short position (sold more than bought)
+negative_position = position[position < 0]
+```
+
+---
+
+## 🔄 Data Processing Pipeline
+
+```
+Polygon Blockchain
+       ↓
+   RPC Query (OrderFilled events)
+       ↓
+   Decode ABI
+       ↓
+orderfilled.parquet (31GB)
+   ├─→ Link market_id via Gamma API
+   ↓
+trades.parquet (32GB)
+   ├─→ Filter + Unified YES perspective
+   ├─→ Keep maker/taker structure
+   ↓
+quant.parquet (21GB)
+   │
+   └─→ Split maker/taker + Unified BUY direction
+       ↓
+   users.parquet (23GB)
+```
+
+---
+
+## 🆚 Comparison: quant vs users
+
+| Feature | quant.parquet | users.parquet |
+|---------|---------------|---------------|
+| **Perspective** | Trade-centric | User-centric |
+| **Records per trade** | 1 | 2 (maker + taker) |
+| **Size** | 21GB | 23GB |
+| **Token normalization** | ✅ YES (token1) | ✅ YES (token1) |
+| **Direction** | Preserved (BUY/SELL) | Unified (BUY only) |
+| **Maker/Taker** | Both preserved | Split into rows |
+| **Sort order** | Original | User → Time |
+| **Use case** | Trading algorithms | User behavior |
+
+---
+
+## 💡 Usage Examples
+
+### Example 1: Load Quant Data for Backtesting
+```python
+import pandas as pd
+
+quant = pd.read_parquet('quant.parquet')
+
+# All trades in YES token perspective
+# Price represents YES probability
+market_data = quant[quant['market_id'] == 'specific_market']
+
+# Calculate returns
+market_data = market_data.sort_values('timestamp')
+market_data['returns'] = market_data['price'].pct_change()
+```
+
+### Example 2: Analyze User Trading Patterns
+```python
+users = pd.read_parquet('users.parquet')
+
+# Get one user's trading history
+user_trades = users[users['user'] == '0x123...']
+
+# Calculate net position (positive = long, negative = short)
+net_position = user_trades.groupby('market_id')['token_amount'].sum()
+
+# Identify strategy
+if (user_trades['role'] == 'maker').mean() > 0.7:
+    strategy = 'Market Maker'
+elif net_position.abs().mean() < 1000:
+    strategy = 'Scalper'
+else:
+    strategy = 'Position Trader'
+```
+
+---
+
+## 📊 Data Quality
+
+### Completeness
+- ✅ All OrderFilled events since contract deployment
+- ✅ No missing blocks (failed fetches are retried)
+- ✅ All blockchain fields preserved
+
+### Cleaning Rules
+- ❌ Removed: Contract-to-contract trades
+- ❌ Removed: Trades with NaN prices
+- ✅ Kept: All user-to-user trades
+- ✅ Kept: All fee information
+
+### Token Normalization
+- All `quant.parquet` and `users.parquet` prices represent **YES token probability**
+- Original token2 prices are flipped: `price_yes = 1 - price_no`
+- Consistent for time-series and cross-market analysis
+
+---
+
+## 🔗 Relationships
+
+```
+markets.parquet
+    ↓ (market_id)
+trades.parquet
+    ↓ (filter + normalize)
+quant.parquet (trade-level)
+users.parquet (user-level)
+```
+
+---
+
+## ⚖️ License
+
+MIT License - Free for commercial and research use
+
+---
+
+## 📞 Questions?
+
+- GitHub Issues: [polymarket-data](https://github.com/SII-WANGZJ/polymarket-data/issues)
+- Dataset: [HuggingFace](https://huggingface.co/datasets/SII-WANGZJ/Polymarket_data)

README.md

@@ -0,0 +1,429 @@
+<div align="center">
+
+<h1>Polymarket Data</h1>
+
+<h3>Complete Data Infrastructure for Polymarket — Fetch, Process, Analyze</h3>
+
+<p style="max-width: 750px; margin: 0 auto;">
+A comprehensive dataset of 1.1 billion trading records from Polymarket, processed into multiple analysis-ready formats. Features cleaned data, unified token perspectives, and user-level transformations — ready for market research, behavioral studies, and quantitative analysis.
+</p>
+
+<p>
+<b>Zhengjie Wang</b><sup>1,2</sup>, <b>Leiyu Chao</b><sup>1,3</sup>, <b>Yu Bao</b><sup>1,4</sup>, <b>Lian Cheng</b><sup>1,3</sup>, <b>Jianhan Liao</b><sup>1,5</sup>, <b>Yikang Li</b><sup>1,†</sup>
+</p>
+
+<p>
+<sup>1</sup>Shanghai Innovation Institute &nbsp;&nbsp; <sup>2</sup>Westlake University &nbsp;&nbsp; <sup>3</sup>Shanghai Jiao Tong University
+<br>
+<sup>4</sup>Harbin Institute of Technology &nbsp;&nbsp; <sup>5</sup>Fudan University
+</p>
+
+<p>
+<sup>†</sup>Corresponding author
+</p>
+
+</div>
+
+<p align="center">
+  <a href="https://huggingface.co/datasets/SII-WANGZJ/Polymarket_data">
+    <img src="https://img.shields.io/badge/Hugging%20Face-Dataset-yellow.svg" alt="HuggingFace Dataset"/>
+  </a>
+  <a href="https://github.com/SII-WANGZJ/Polymarket_data">
+    <img src="https://img.shields.io/badge/GitHub-Code-black.svg?logo=github" alt="GitHub Repository"/>
+  </a>
+  <a href="https://github.com/SII-WANGZJ/Polymarket_data/blob/main/LICENSE">
+    <img src="https://img.shields.io/badge/License-MIT-blue.svg" alt="License"/>
+  </a>
+  <a href="#data-quality">
+    <img src="https://img.shields.io/badge/Data-Verified-green.svg" alt="Data Quality"/>
+  </a>
+</p>
+
+---
+
+## TL;DR
+
+We provide **107GB of historical on-chain trading data** from Polymarket, containing **1.1 billion records** across 268K+ markets. The dataset is directly fetched from Polygon blockchain, fully verified, and ready for analysis. Perfect for market research, behavioral studies, data science projects, and academic research.
+
+## Highlights
+
+- **Complete Blockchain History**: All OrderFilled events from Polymarket's two exchange contracts, with no missing blocks or gaps. Every single trade from the platform's inception is included.
+
+- **Multiple Analysis Perspectives**: 5 carefully curated datasets serving different research needs - from raw blockchain events to user-level behavior analysis, with unified data transformations for easy analysis.
+
+- **Production Ready**: Clean, validated data with proper schema documentation. All trades are verified against blockchain RPC, with market metadata linked and ready to use.
+
+- **Open Source Pipeline**: Fully reproducible data collection process. Our open-source tools allow you to verify, update, or extend the dataset independently.
+
+## Dataset Overview
+
+| File | Size | Records | Description |
+|------|------|---------|-------------|
+| `orderfilled.parquet` | 31GB | 293.3M | Raw blockchain events from OrderFilled logs |
+| `trades.parquet` | 32GB | 293.3M | Processed trades with market metadata linkage |
+| `markets.parquet` | 68MB | 268,706 | Market information and metadata |
+| `quant.parquet` | 21GB | 170.3M | Clean market data with unified YES perspective |
+| `users.parquet` | 23GB | 340.6M | User behavior data split by maker/taker roles |
+
+**Total**: 107GB, 1.1 billion records
+
+## Use Cases
+
+### Market Research & Analysis
+- Study prediction market dynamics and price discovery mechanisms
+- Analyze market efficiency and information aggregation
+- Research crowd wisdom and forecasting accuracy
+
+### Behavioral Studies
+- Track individual user trading patterns and decision-making
+- Study market participant behavior under different conditions
+- Analyze risk preferences and trading strategies
+
+### Data Science & Machine Learning
+- Train models for price prediction and market forecasting
+- Feature engineering for time-series analysis
+- Develop algorithms for market analysis
+
+### Academic Research
+- Economics and finance research on prediction markets
+- Social science studies on collective intelligence
+- Computer science research on blockchain data analysis
+
+## Quick Start
+
+### Installation
+
+```bash
+# Using pip
+pip install pandas pyarrow
+
+# Optional: for faster parquet reading
+pip install fastparquet
+```
+
+### Load Data with Pandas
+
+```python
+import pandas as pd
+
+# Load clean market data
+df = pd.read_parquet('quant.parquet')
+print(f"Total trades: {len(df):,}")
+
+# Load user behavior data
+users = pd.read_parquet('users.parquet')
+print(f"Total user actions: {len(users):,}")
+
+# Load market metadata
+markets = pd.read_parquet('markets.parquet')
+print(f"Total markets: {len(markets):,}")
+```
+
+### Load from HuggingFace Datasets
+
+```python
+from datasets import load_dataset
+
+# Load specific file
+dataset = load_dataset(
+    "SII-WANGZJ/Polymarket_data",
+    data_files="quant.parquet"
+)
+
+# Load multiple files
+dataset = load_dataset(
+    "SII-WANGZJ/Polymarket_data",
+    data_files=["quant.parquet", "markets.parquet"]
+)
+```
+
+### Download Specific Files
+
+```bash
+# Download using HuggingFace CLI
+pip install huggingface_hub
+
+# Download a specific file
+hf download SII-WANGZJ/Polymarket_data quant.parquet --repo-type dataset
+
+# Download all files
+hf download SII-WANGZJ/Polymarket_data --repo-type dataset
+```
+
+## Data Structure
+
+### quant.parquet - Clean Market Data
+
+Filtered and normalized trade data with unified token perspective (YES token).
+
+**Key Features:**
+- Unified perspective: All trades normalized to YES token (token1)
+- Clean data: Contract trades filtered out, only real user trades
+- Complete information: Maker/taker roles preserved
+- Best for: Market analysis, price studies, time-series forecasting
+
+**Schema:**
+```python
+{
+    'transaction_hash': str,      # Blockchain transaction hash
+    'block_number': int,          # Block number
+    'datetime': datetime,         # Transaction timestamp
+    'market_id': str,             # Market identifier
+    'maker': str,                 # Maker wallet address
+    'taker': str,                 # Taker wallet address
+    'token_amount': float,        # Amount of tokens traded
+    'usd_amount': float,          # USD value
+    'price': float,               # Trade price (0-1)
+}
+```
+
+### users.parquet - User Behavior Data
+
+Split maker/taker records with unified buy direction for user analysis.
+
+**Key Features:**
+- Split records: Each trade becomes 2 records (one maker, one taker)
+- Unified direction: All converted to BUY (negative amounts = selling)
+- User sorted: Ordered by user for trajectory analysis
+- Best for: User profiling, PnL calculation, wallet analysis
+
+**Schema:**
+```python
+{
+    'transaction_hash': str,      # Transaction hash
+    'block_number': int,          # Block number
+    'datetime': datetime,         # Timestamp
+    'market_id': str,             # Market identifier
+    'user': str,                  # User wallet address
+    'role': str,                  # 'maker' or 'taker'
+    'token_amount': float,        # Signed amount (+ buy, - sell)
+    'usd_amount': float,          # USD value
+    'price': float,               # Trade price
+}
+```
+
+### markets.parquet - Market Metadata
+
+Market information and outcome token details.
+
+**Best for:** Linking trades to market context, filtering by market attributes
+
+### trades.parquet - Processed Blockchain Data
+
+Raw OrderFilled events with market linkage but no transformations.
+
+**Best for:** Custom analysis requiring original blockchain data
+
+### orderfilled.parquet - Raw Blockchain Events
+
+Unprocessed OrderFilled events directly from blockchain logs.
+
+**Best for:** Blockchain research, verification, custom processing pipelines
+
+## Example Analysis
+
+### 1. Calculate Market Statistics
+
+```python
+import pandas as pd
+
+df = pd.read_parquet('quant.parquet')
+
+# Market-level statistics
+market_stats = df.groupby('market_id').agg({
+    'usd_amount': ['sum', 'mean'],     # Total volume and average trade size
+    'price': ['mean', 'std', 'min', 'max'],  # Price statistics
+    'transaction_hash': 'count'         # Number of trades
+}).round(4)
+
+print(market_stats.head())
+```
+
+### 2. Track Price Evolution
+
+```python
+import pandas as pd
+import matplotlib.pyplot as plt
+
+df = pd.read_parquet('quant.parquet')
+df['datetime'] = pd.to_datetime(df['datetime'])
+
+# Select a specific market
+market_id = 'your-market-id'
+market_data = df[df['market_id'] == market_id].sort_values('datetime')
+
+# Plot price over time
+plt.figure(figsize=(12, 6))
+plt.plot(market_data['datetime'], market_data['price'])
+plt.title(f'Price Evolution - Market {market_id}')
+plt.xlabel('Date')
+plt.ylabel('Price')
+plt.show()
+```
+
+### 3. Analyze User Behavior
+
+```python
+import pandas as pd
+
+df = pd.read_parquet('users.parquet')
+
+# Calculate net position per user per market
+user_positions = df.groupby(['user', 'market_id']).agg({
+    'token_amount': 'sum',          # Net position (positive = long, negative = short)
+    'usd_amount': 'sum',            # Total USD traded
+    'transaction_hash': 'count'     # Number of trades
+}).reset_index()
+
+# Find most active users
+active_users = user_positions.groupby('user').agg({
+    'market_id': 'count',           # Number of markets traded
+    'usd_amount': 'sum'             # Total volume
+}).sort_values('usd_amount', ascending=False)
+
+print(active_users.head(10))
+```
+
+### 4. Market Volume Analysis
+
+```python
+import pandas as pd
+
+df = pd.read_parquet('quant.parquet')
+markets = pd.read_parquet('markets.parquet')
+
+# Join with market metadata
+df = df.merge(markets[['market_id', 'question']], on='market_id', how='left')
+
+# Top markets by volume
+top_markets = df.groupby(['market_id', 'question']).agg({
+    'usd_amount': 'sum'
+}).sort_values('usd_amount', ascending=False).head(20)
+
+print(top_markets)
+```
+
+## Data Processing Pipeline
+
+```
+Polygon Blockchain (RPC)
+         ↓
+  orderfilled.parquet (Raw events)
+         ↓
+  trades.parquet (+ Market linkage)
+         ↓
+         ├─→ quant.parquet (Trade-level, unified YES perspective)
+         │   └─→ Filter contracts + Normalize tokens
+         │
+         └─→ users.parquet (User-level, split maker/taker)
+             └─→ Split records + Unified BUY direction
+```
+
+**Key Transformations:**
+
+1. **quant.parquet**:
+   - Filter out contract trades (keep only user trades)
+   - Normalize all trades to YES token perspective
+   - Preserve maker/taker information
+   - Result: 170.3M records (from 293.3M)
+
+2. **users.parquet**:
+   - Split each trade into 2 records (maker + taker)
+   - Convert all to BUY direction (signed amounts)
+   - Sort by user for easy querying
+   - Result: 340.6M records (from 293.3M × 2, some filtered)
+
+## Documentation
+
+- **[DATA_DESCRIPTION.md](DATA_DESCRIPTION.md)** - Comprehensive documentation
+  - Detailed schema for all 5 files
+  - Data cleaning and transformation process
+  - Usage examples and best practices
+  - Comparison between different files
+
+## Data Quality
+
+- **Complete History**: No missing blocks or gaps in blockchain data
+- **Verified Sources**: All OrderFilled events from 2 official exchange contracts
+- **Blockchain Verified**: Cross-checked against Polygon RPC nodes
+- **Regular Updates**: Automated daily pipeline for fresh data
+- **Open Source**: Fully reproducible collection process
+
+**Contracts Tracked:**
+- Exchange Contract 1: `0x4bFb41d5B3570DeFd03C39a9A4D8dE6Bd8B8982E`
+- Exchange Contract 2: `0xC5d563A36AE78145C45a50134d48A1215220f80a`
+
+## Collection Tools
+
+Data collected using our open-source toolkit: [polymarket-data](https://github.com/SII-WANGZJ/Polymarket_data)
+
+**Features:**
+- Direct blockchain RPC integration
+- Efficient batch processing
+- Automatic retry and error handling
+- Data validation and verification
+
+## Dataset Statistics
+
+**Last Updated**: 2026-01-01
+
+**Coverage**:
+- Time Range: [Polymarket inception] to [Latest update]
+- Total Markets: 268,706
+- Total Trades: 293.3 million
+- Total Volume: $[To be calculated] billion
+- Unique Users: [To be calculated]
+
+**Data Freshness**: Updated daily via automated pipeline
+
+## Contributing
+
+We welcome contributions to improve the dataset and tools:
+
+1. **Report Issues**: Found data quality issues? [Open an issue](https://github.com/SII-WANGZJ/Polymarket_data/issues)
+2. **Suggest Features**: Ideas for new data transformations? Let us know!
+3. **Contribute Code**: Improve our collection pipeline via pull requests
+
+## License
+
+MIT License - Free for commercial and research use.
+
+See [LICENSE](LICENSE) file for details.
+
+## Contact & Support
+
+- **Email**: [wangzhengjie@sii.edu.cn](mailto:wangzhengjie@sii.edu.cn)
+- **Issues**: [GitHub Issues](https://github.com/SII-WANGZJ/Polymarket_data/issues)
+- **Dataset**: [HuggingFace](https://huggingface.co/datasets/SII-WANGZJ/Polymarket_data)
+- **Code**: [GitHub Repository](https://github.com/SII-WANGZJ/Polymarket_data)
+
+## Citation
+
+If you use this dataset in your research, please cite:
+
+```bibtex
+@misc{polymarket_data_2026,
+  title={Polymarket Data: Complete Data Infrastructure for Polymarket},
+  author={Wang, Zhengjie and Chao, Leiyu and Bao, Yu and Cheng, Lian and Liao, Jianhan and Li, Yikang},
+  year={2026},
+  howpublished={\url{https://huggingface.co/datasets/SII-WANGZJ/Polymarket_data}},
+  note={A comprehensive dataset and toolkit for Polymarket prediction markets}
+}
+```
+
+## Acknowledgments
+
+- **Polymarket** for building the leading prediction market platform
+- **Polygon** for providing reliable blockchain infrastructure
+- **HuggingFace** for hosting and distributing large datasets
+- The open-source community for tools and libraries
+
+---
+
+<div align="center">
+
+**Built for the research and data science community**
+
+[HuggingFace](https://huggingface.co/datasets/SII-WANGZJ/Polymarket_data) • [GitHub](https://github.com/SII-WANGZJ/Polymarket_data) • [Documentation](DATA_DESCRIPTION.md)
+
+</div>

markets.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e4533e08cebba2bc22a4c089df7b23e96451cacafed693a4ec9d0a48b230921
+size 70941035

orderfilled.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:481809ee64543da09ed84f282c4f2805c7b4d1ff8d1cefd631a6df7d5dd6c055
+size 33518172323

quant.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46079dbe3b4aa16dbfd4535b10846529fcd189dbdc4ebd38d1e7a391eb6b9bba
+size 22859967175

trades.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17386c6ebd146f756548b4be05cac3d8ee511a7f5e4e720340b22edd04412ca4
+size 34709564052

users.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:766b1925196f5003f4f1e2868f6c8ef825f87da023fca5b01e1b11afcc4e45c1
+size 28929603335