Upload 49 files

ADA_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8da4c283fcdc958be0de2ad578ce39304a8702455e1cd5cbefeaf66433e6577
+size 651636

APT_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ee59677be1fad8881bed53fe59f47832cb1c8bc74171248c2d2fde2bbaf63d2
+size 662669

ARB_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c9edd185a0744552e5342ab0058f5dffd4c23e9cac627c849995c1cde6ebb5b
+size 663904

ATOM_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9e91ec95941404575dd56fa8555913c90e4d2a7b88bd600abca9c45d1cf1959
+size 679264

AVAX_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f957a4d7fc2a8238219a21d3eeb638b231c1db17d8b85c43526c1967482da953
+size 657228

BNB_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:390c6a576b090f6a2a398cd25aa867b74141d764edbbe6dbc5648560daf0b05c
+size 622035

BTC_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88ac5d30a8edc4ceb3f1c4218caca9022e3bca99fe6a708efdb71701cc93bcf8
+size 631398

DOGE_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e48cb5d2be6c1d35f0efcee4ba10eba9ce8015ecffa5a855ac77b88b724994ea
+size 623069

DOT_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:765167791985ebd9f981854e3fdda0d20552a3c1cfefdddaede1c4977251c999
+size 637978

ETC_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63bc43b1e627130ec37af409a6d16ba3dd4e15eab83eec20cab0e421fc673d37
+size 659962

ETH_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fe572cd0d84a61b6ea4239748bc9064d059224046197ce2450db9b9ca1824ed
+size 601889

FIL_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8789217d187888c6c83909492001101de7cedea52227822d96bcbc8eaaa2719e
+size 689707

INJ_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3f0c2ba1aa855173ce6c22a497dcc3f9021d098171039aea45bc7b9e18ca300
+size 654992

LINK_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a8245e332ab7e038e54095a46b7d43957c13306755a6959d1fa43d41aeff59f
+size 642559

LTC_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6339525176ea84335f29e74a021cf7f5738a4ad19a695990dacb2f4cfe52ae84
+size 627743

NEAR_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa5d7918409b14e52cf547dd008abf5da14eb4ee023c991c68353986a146dbf3
+size 633057

OP_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30e277b49634df7e5d97f962f3ee63b7f9eff766cdc5162130e04c56a303cf3b
+size 687502

SEI_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf272039ea9d397e6ac73058d107f4ca81bf0792640c0d5d601e2627e629f1ba
+size 666769

SOL_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a8a6f2b85e1582fd4309020cd6eb609e5d39fb0b1ba01fbdf23f5b3965c7bc9
+size 599349

SUI_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c905c71f55d0cbed9547fcaa6b4ca3639a5ef022b54a205eea788ab04f2e0d78
+size 672079

TIA_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d353f6baec5e40fa083e41058a4fe328f18193dc28b0e3394a3d81bc1e0b6206
+size 686490

WIF_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50c3119b538df16a97caec86618e2734267ca91ed3ee328e765febb3408a0efc
+size 650579

XLM_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28b9968a81f92008a9abd57e8dc346007d3b4d3b7a2334915548fb4a8935cb27
+size 648731

XRP_sample_7d.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:113ef3290f5aa7ced94566329726710d1213adcc2a77f0ada528c77a08153e95
+size 554758

imbalance_labs_eda_starter.ipynb

@@ -0,0 +1,313 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# 📊 Imbalance Labs — EDA Starter Kit\n",
+    "\n",
+    "This notebook walks you through loading, exploring, and visualizing the Imbalance Labs orderbook dataset.\n",
+    "\n",
+    "**Dataset specs:**\n",
+    "- 24 crypto instruments (BTC, ETH, SOL, etc.)\n",
+    "- 5-minute aggregated bars, 12+ months history\n",
+    "- 47 columns per row: OHLC + 10-level depth (bid/ask volumes + distances)\n",
+    "\n",
+    "---"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 1. Setup & Load Data"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import numpy as np\n",
+    "import matplotlib.pyplot as plt\n",
+    "import matplotlib.dates as mdates\n",
+    "\n",
+    "plt.style.use('dark_background')\n",
+    "plt.rcParams['figure.figsize'] = (14, 6)\n",
+    "plt.rcParams['font.size'] = 11\n",
+    "\n",
+    "ACCENT = '#00FF88'\n",
+    "\n",
+    "# ── Load a single instrument ──\n",
+    "# Replace with your file path\n",
+    "df = pd.read_csv('BTC_5m_depth10_derived.csv.gz')\n",
+    "df['timestamp_utc'] = pd.to_datetime(df['timestamp_utc'])\n",
+    "df = df.set_index('timestamp_utc').sort_index()\n",
+    "\n",
+    "print(f'Instrument: {df[\"instrument_symbol\"].iloc[0]}')\n",
+    "print(f'Rows: {len(df):,}')\n",
+    "print(f'Date range: {df.index.min()} → {df.index.max()}')\n",
+    "print(f'Columns: {len(df.columns)}')\n",
+    "df.head()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 2. Price Overview"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(14, 8), gridspec_kw={'height_ratios': [3, 1]}, sharex=True)\n",
+    "\n",
+    "ax1.plot(df.index, df['close_price'], color=ACCENT, linewidth=0.7, alpha=0.9)\n",
+    "ax1.fill_between(df.index, df['low_price'], df['high_price'], alpha=0.1, color=ACCENT)\n",
+    "ax1.set_ylabel('Price (USDT)')\n",
+    "ax1.set_title(f'{df[\"instrument_symbol\"].iloc[0]} — Close Price', fontweight='bold')\n",
+    "ax1.grid(alpha=0.15)\n",
+    "\n",
+    "ax2.bar(df.index, df['interval_traded_volume'], width=0.003, color=ACCENT, alpha=0.5)\n",
+    "ax2.set_ylabel('Volume')\n",
+    "ax2.set_xlabel('Date')\n",
+    "ax2.grid(alpha=0.15)\n",
+    "\n",
+    "ax1.xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m'))\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 3. Orderbook Depth Profile\n",
+    "\n",
+    "Visualize how liquidity is distributed across the 10 depth levels."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Average depth profile across entire dataset\n",
+    "bid_cols = [f'bid_volume_level_{i}' for i in range(1, 11)]\n",
+    "ask_cols = [f'ask_volume_level_{i}' for i in range(1, 11)]\n",
+    "\n",
+    "avg_bid = df[bid_cols].mean().values\n",
+    "avg_ask = df[ask_cols].mean().values\n",
+    "\n",
+    "levels = np.arange(1, 11)\n",
+    "\n",
+    "fig, ax = plt.subplots(figsize=(10, 6))\n",
+    "ax.barh(levels - 0.2, avg_bid, height=0.35, color='#00FF88', alpha=0.8, label='Bid (Buy Wall)')\n",
+    "ax.barh(levels + 0.2, avg_ask, height=0.35, color='#FF4444', alpha=0.8, label='Ask (Sell Wall)')\n",
+    "ax.set_ylabel('Depth Level')\n",
+    "ax.set_xlabel('Average Cumulative Volume')\n",
+    "ax.set_title('Orderbook Depth Profile — Average Liquidity by Level', fontweight='bold')\n",
+    "ax.set_yticks(levels)\n",
+    "ax.legend()\n",
+    "ax.grid(alpha=0.15, axis='x')\n",
+    "ax.invert_yaxis()\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 4. Bid-Ask Imbalance\n",
+    "\n",
+    "The **imbalance ratio** measures the relative pressure between buyers and sellers at each level.\n",
+    "\n",
+    "$$\\text{Imbalance}_k = \\frac{\\text{Bid}_k - \\text{Ask}_k}{\\text{Bid}_k + \\text{Ask}_k}$$\n",
+    "\n",
+    "A value close to **+1** means heavy buy-side pressure; **−1** means sell-side dominance."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Compute imbalance for level 1 (tightest)\n",
+    "df['imbalance_L1'] = (\n",
+    "    (df['bid_volume_level_1'] - df['ask_volume_level_1']) /\n",
+    "    (df['bid_volume_level_1'] + df['ask_volume_level_1'])\n",
+    ")\n",
+    "\n",
+    "# Rolling smoothed version\n",
+    "df['imbalance_L1_smooth'] = df['imbalance_L1'].rolling(60).mean()  # 5h rolling avg\n",
+    "\n",
+    "# ── Plot ──\n",
+    "fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(14, 8), sharex=True)\n",
+    "\n",
+    "ax1.plot(df.index, df['close_price'], color=ACCENT, linewidth=0.6)\n",
+    "ax1.set_ylabel('Price')\n",
+    "ax1.set_title('Price vs Level-1 Bid-Ask Imbalance', fontweight='bold')\n",
+    "ax1.grid(alpha=0.15)\n",
+    "\n",
+    "ax2.fill_between(df.index, df['imbalance_L1_smooth'],\n",
+    "                 where=df['imbalance_L1_smooth'] > 0, color='#00FF88', alpha=0.5, label='Buy pressure')\n",
+    "ax2.fill_between(df.index, df['imbalance_L1_smooth'],\n",
+    "                 where=df['imbalance_L1_smooth'] < 0, color='#FF4444', alpha=0.5, label='Sell pressure')\n",
+    "ax2.axhline(0, color='white', linewidth=0.5, alpha=0.3)\n",
+    "ax2.set_ylabel('Imbalance (L1)')\n",
+    "ax2.set_ylim(-0.5, 0.5)\n",
+    "ax2.legend(loc='upper right')\n",
+    "ax2.grid(alpha=0.15)\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 5. Depth Distance Heatmap\n",
+    "\n",
+    "How far is the liquidity from mid-price at each level? Tighter spread = more liquid market."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "bid_dist_cols = [f'bid_distance_level_{i}' for i in range(1, 11)]\n",
+    "ask_dist_cols = [f'ask_distance_level_{i}' for i in range(1, 11)]\n",
+    "\n",
+    "# Resample to daily for cleaner heatmap\n",
+    "daily_bid_dist = df[bid_dist_cols].resample('1D').mean()\n",
+    "daily_ask_dist = df[ask_dist_cols].resample('1D').mean()\n",
+    "\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(16, 8))\n",
+    "\n",
+    "im1 = ax1.imshow(daily_bid_dist.T, aspect='auto', cmap='YlGn', interpolation='nearest')\n",
+    "ax1.set_title('Bid Distance from Mid-Price (bps)', fontweight='bold')\n",
+    "ax1.set_ylabel('Depth Level')\n",
+    "ax1.set_yticks(range(10))\n",
+    "ax1.set_yticklabels([f'L{i}' for i in range(1, 11)])\n",
+    "ax1.set_xlabel('Day')\n",
+    "plt.colorbar(im1, ax=ax1, label='bps')\n",
+    "\n",
+    "im2 = ax2.imshow(daily_ask_dist.T, aspect='auto', cmap='YlOrRd', interpolation='nearest')\n",
+    "ax2.set_title('Ask Distance from Mid-Price (bps)', fontweight='bold')\n",
+    "ax2.set_ylabel('Depth Level')\n",
+    "ax2.set_yticks(range(10))\n",
+    "ax2.set_yticklabels([f'L{i}' for i in range(1, 11)])\n",
+    "ax2.set_xlabel('Day')\n",
+    "plt.colorbar(im2, ax=ax2, label='bps')\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 6. Multi-Instrument Comparison\n",
+    "\n",
+    "Load all instruments and compare their average imbalance."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import glob\n",
+    "\n",
+    "files = sorted(glob.glob('*_5m_depth10_derived.csv.gz'))\n",
+    "print(f'Found {len(files)} instruments')\n",
+    "\n",
+    "stats = []\n",
+    "for f in files:\n",
+    "    d = pd.read_csv(f)\n",
+    "    sym = d['instrument_symbol'].iloc[0]\n",
+    "    imb = (d['bid_volume_level_1'] - d['ask_volume_level_1']) / (d['bid_volume_level_1'] + d['ask_volume_level_1'])\n",
+    "    total_vol = d['interval_traded_volume'].sum()\n",
+    "    stats.append({\n",
+    "        'instrument': sym,\n",
+    "        'rows': len(d),\n",
+    "        'avg_imbalance_L1': imb.mean(),\n",
+    "        'std_imbalance_L1': imb.std(),\n",
+    "        'total_volume': total_vol,\n",
+    "        'avg_spread_L1_bps': (d['ask_distance_level_1'] + d['bid_distance_level_1']).mean(),\n",
+    "    })\n",
+    "\n",
+    "comparison = pd.DataFrame(stats).set_index('instrument').sort_values('total_volume', ascending=False)\n",
+    "comparison"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig, ax = plt.subplots(figsize=(14, 6))\n",
+    "colors = [ACCENT if v > 0 else '#FF4444' for v in comparison['avg_imbalance_L1']]\n",
+    "ax.bar(comparison.index, comparison['avg_imbalance_L1'], color=colors, alpha=0.8)\n",
+    "ax.axhline(0, color='white', linewidth=0.5, alpha=0.3)\n",
+    "ax.set_ylabel('Average L1 Bid-Ask Imbalance')\n",
+    "ax.set_title('Cross-Instrument Imbalance Comparison', fontweight='bold')\n",
+    "ax.grid(alpha=0.15, axis='y')\n",
+    "plt.xticks(rotation=45)\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 7. Feature Engineering Ideas\n",
+    "\n",
+    "Here are some features you can derive from this dataset for ML models:\n",
+    "\n",
+    "| Feature | Formula | Use Case |\n",
+    "|---|---|---|\n",
+    "| **Imbalance Ratio (L1-L10)** | `(bid_vol - ask_vol) / (bid_vol + ask_vol)` | Directional signal |\n",
+    "| **Depth-Weighted Imbalance** | Σ `imbalance_k × (1/k)` for k=1..10 | Weighted signal favoring top-of-book |\n",
+    "| **Total Depth** | Σ `bid_vol + ask_vol` for k=1..10 | Liquidity regime detection |\n",
+    "| **Depth Slope** | Linear regression slope of volume vs level | Wall detection |\n",
+    "| **Spread Momentum** | `diff(bid_distance_L1 + ask_distance_L1)` | Spread widening/tightening |\n",
+    "| **Volume-Distance Ratio** | `volume_Lk / distance_Lk` | Concentration score |\n",
+    "| **Cross-Level Divergence** | `imbalance_L1 - imbalance_L5` | Near vs far pressure gap |\n",
+    "\n",
+    "---\n",
+    "\n",
+    "🔗 **imbalancelabs.com** — Full dataset: 24 instruments, 12+ months, 47 columns/row."
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "name": "python",
+   "version": "3.11.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}