Add GitHub link, analysis notebook, set public

HUGGINGFACE_README.md

@@ -0,0 +1,114 @@
+---
+license: cc-by-4.0
+task_categories:
+  - graph-ml
+  - feature-extraction
+language:
+  - en
+tags:
+  - steam
+  - games
+  - network-analysis
+  - co-review
+  - graph-data
+  - gaming
+  - social-network
+pretty_name: Steam Co-Review Network
+size_categories:
+  - 1M<n<10M
+dataset_info:
+  features:
+    - name: nodes
+      dtype: list
+    - name: links
+      dtype: list
+    - name: meta
+      dtype: dict
+---
+
+# Steam Co-Review Network
+
+Two Steam games share an edge when multiple users reviewed both. Built from 128 million user reviews across 80,000 games (2012-2024), this dataset maps how the Steam catalog is connected through player overlap.
+
+## Files
+
+### steam_network_full.json
+
+The complete co-review graph with minimal filtering:
+
+- 48,362 game nodes (every game with 10+ reviews)
+- 33,041,298 weighted edges (2+ shared reviewers per pair)
+- Per-node cap of 50 neighbors (densest connections preserved)
+- Edge weights range from 2 to 420,410 shared reviewers
+
+**Node format:**
+```json
+{
+  "id": "620",
+  "title": "Portal 2",
+  "year": "2011",
+  "rating": "Overwhelmingly Positive",
+  "ratio": 97,
+  "reviews": 263842,
+  "price": 9.99
+}
+```
+
+**Link format:**
+```json
+{"source": 0, "target": 42, "weight": 1523}
+```
+
+Source and target are indices into the nodes array. Weight is the number of users who reviewed both games.
+
+### steam_all_2005.json
+
+82,928 games released 2005-2025. Packed as arrays for compact JSON:
+
+```
+[name, year, ratio, reviews, price, ratingIdx, genreIdxs, tagIdxs, developer]
+ [0]    [1]   [2]    [3]      [4]    [5]        [6]        [7]       [8]
+```
+
+Genres and tags are stored as index arrays referencing top-level `genres[]` and `tags[]` lookup tables in the same file.
+
+### steam_force_layout.json
+
+Genre-aware pre-computed layout positions for the top ~9K nodes, clustered by primary genre with hub games as anchors. Use as warm-start coordinates for force-directed visualization.
+
+## Sources
+
+- **Game metadata**: [FronkonGames Steam Games Dataset](https://huggingface.co/datasets/FronkonGames/steam-games-dataset) — Jan 2026 snapshot, 122K games
+- **User reviews**: [artermiloff Steam Reviews 2024](https://www.kaggle.com/datasets/artermiloff/steam-games-reviews-2024) — 128M reviews across 80K games, one CSV per game, 2012-June 2024
+
+## Pipeline
+
+1. Load game metadata from FronkonGames enriched CSV
+2. Scan 30K+ per-game review CSVs, extract steamid-to-game mappings
+3. For each user who reviewed 2+ games, generate all game pairs
+4. Count shared reviewers per pair to produce edge weights
+5. Filter: minimum 2 shared reviewers (no neighbor cap)
+
+Full pipeline: [github.com/lukeslp/steam-network-data](https://github.com/lukeslp/steam-network-data)
+
+## Use Cases
+
+- **Graph ML**: Node classification (predict genre/rating from network position), link prediction, community detection
+- **Recommendation systems**: Games connected by high edge weights share audiences
+- **Market analysis**: Which genres cluster together? Where are the gaps?
+- **Visualization**: Force-directed layouts, chord diagrams, genre timelines of the Steam ecosystem
+
+## Live Visualization
+
+[dr.eamer.dev/datavis/interactive/steam/](https://dr.eamer.dev/datavis/interactive/steam/)
+
+Four interactive Canvas-rendered views: universe scatter, chord diagram, force-directed network, and genre timeline.
+
+## Distribution
+
+- **GitHub**: [lukeslp/steam-network-data](https://github.com/lukeslp/steam-network-data)
+- **Kaggle**: [lucassteuber/steam-universe-network](https://www.kaggle.com/datasets/lucassteuber/steam-universe-network)
+
+## Author
+
+Luke Steuber — [lukesteuber.com](https://lukesteuber.com) — [@lukesteuber.com on Bluesky](https://bsky.app/profile/lukesteuber.com)

README.md

@@ -1,109 +1,41 @@
----
-license: cc-by-4.0
-task_categories:
-  - graph-ml
-  - feature-extraction
-language:
-  - en
-tags:
-  - steam
-  - games
-  - network-analysis
-  - co-review
-  - graph-data
-  - gaming
-  - social-network
-pretty_name: Steam Co-Review Network
-size_categories:
-  - 1M<n<10M
-dataset_info:
-  features:
-    - name: nodes
-      dtype: list
-    - name: links
-      dtype: list
-    - name: meta
-      dtype: dict
----
-
 # Steam Co-Review Network
 
-Two Steam games share an edge when multiple users reviewed both. Built from 128 million user reviews across 80,000 games (2012-2024), this dataset maps how the Steam catalog is connected through player overlap.
-
-## Files
-
-### steam_network_full.json
-
-The complete co-review graph with minimal filtering:
-
-- 48,362 game nodes (every game with 10+ reviews)
-- 33,041,298 weighted edges (2+ shared reviewers per pair)
-- Per-node cap of 50 neighbors (densest connections preserved)
-- Edge weights range from 2 to 694,377 shared reviewers
-
-**Node format:**
-```json
-{
-  "id": "620",
-  "title": "Portal 2",
-  "year": "2011",
-  "rating": "Overwhelmingly Positive",
-  "ratio": 97,
-  "reviews": 263842,
-  "price": 9.99
-}
-```
-
-**Link format:**
-```json
-{"source": 0, "target": 42, "weight": 1523}
-```
-
-Source and target are indices into the nodes array. Weight is the number of users who reviewed both games.
+Graph dataset connecting 82,000+ Steam games through shared user reviews. Two games share an edge when multiple users reviewed both.
 
-### steam_all_2005.json
+Built from 128 million user reviews across 80,000 games (2012-2024), sourced from the [artermiloff Steam Reviews 2024](https://www.kaggle.com/datasets/artermiloff/steam-games-reviews-2024) dataset and [FronkonGames](https://huggingface.co/datasets/FronkonGames/steam-games-dataset) game metadata.
 
-82,928 games released 2005-2025. Packed as arrays for compact JSON:
-
-```
-[name, year, ratio, reviews, price, ratingIdx, genreIdxs, tagIdxs, developer]
- [0]    [1]   [2]    [3]      [4]    [5]        [6]        [7]       [8]
-```
-
-Genres and tags are stored as index arrays referencing top-level `genres[]` and `tags[]` lookup tables in the same file.
+## Files
 
-### steam_force_layout.json
+| File | Description | Size |
+|------|-------------|------|
+| `steam_network_full.json` | Full co-review graph (48K nodes, 33M weighted edges) | 1.4 GB |
+| `steam_all_2005.json` | 82,928 game catalog with genres, tags, ratings | 6.2 MB |
+| `steam_force_layout.json` | Genre-aware pre-computed layout positions (9K nodes) | 252 KB |
 
-Genre-aware pre-computed layout positions for the top ~9K nodes, clustered by primary genre with hub games as anchors. Use as warm-start coordinates for force-directed visualization.
+## Live Visualization
 
-## Sources
+**[dr.eamer.dev/datavis/interactive/steam/](https://dr.eamer.dev/datavis/interactive/steam/)**
 
-- **Game metadata**: [FronkonGames Steam Games Dataset](https://huggingface.co/datasets/FronkonGames/steam-games-dataset) — Jan 2026 snapshot, 122K games
-- **User reviews**: [artermiloff Steam Reviews 2024](https://www.kaggle.com/datasets/artermiloff/steam-games-reviews-2024) — 128M reviews across 80K games, one CSV per game, 2012-June 2024
+Four interactive Canvas-rendered views: universe scatter plot, chord diagram, force-directed network, and genre timeline.
 
 ## Pipeline
 
-1. Load game metadata from FronkonGames enriched CSV
-2. Scan 30K+ per-game review CSVs, extract steamid-to-game mappings
-3. For each user who reviewed 2+ games, generate all game pairs
-4. Count shared reviewers per pair to produce edge weights
-5. Filter: minimum 2 shared reviewers (no neighbor cap)
-
-Full pipeline: [github.com/lukeslp/steam-game-network](https://github.com/lukeslp/steam-game-network)
-
-## Use Cases
+```bash
+python3 build_network_full.py   # Scan 128M reviews, build co-review edges
+python3 enrich_data.py          # Build game catalog from FronkonGames CSV
+python3 compute_layout.py       # Genre-aware force layout (needs networkx)
+```
 
-- **Graph ML**: Node classification (predict genre/rating from network position), link prediction, community detection
-- **Recommendation systems**: Games connected by high edge weights share audiences
-- **Market analysis**: Which genres cluster together? Where are the gaps?
-- **Visualization**: Force-directed layouts, chord diagrams, genre timelines of the Steam ecosystem
+## Platforms
 
-## Live Visualization
+- **GitHub**: [lukeslp/steam-network-data](https://github.com/lukeslp/steam-network-data)
+- **Kaggle**: [lucassteuber/steam-universe-network](https://www.kaggle.com/datasets/lucassteuber/steam-universe-network)
+- **HuggingFace**: [lukeslp/steam-co-review-network](https://huggingface.co/datasets/lukeslp/steam-co-review-network)
 
-[dr.eamer.dev/datavis/interactive/steam/](https://dr.eamer.dev/datavis/interactive/steam/)
+## License
 
-Four interactive Canvas-rendered views: universe scatter, chord diagram, force-directed network, and genre timeline.
+CC-BY-4.0
 
 ## Author
 
-Luke Steuber — [lukesteuber.com](https://lukesteuber.com) — [@lukesteuber.com on Bluesky](https://bsky.app/profile/lukesteuber.com)
+Luke Steuber — [lukesteuber.com](https://lukesteuber.com)

dataset-metadata.json

@@ -0,0 +1,9 @@
+{
+  "title": "Steam Co-Review Network (82K Games)",
+  "id": "lucassteuber/steam-universe-network",
+  "subtitle": "82K games connected by 128M shared user reviews",
+  "description": "# Steam Co-Review Network\n\nTwo Steam games share an edge when multiple users reviewed both. Built from 128 million user reviews across 80,000 games (2012-2024), this dataset maps how the Steam catalog is connected through player overlap.\n\n## What's Inside\n\n### steam_network_full.json (1.4 GB)\nThe complete co-review graph — 48,362 nodes, 33 million weighted edges:\n- Every game with 10+ reviews included as a node\n- Edges kept when 2+ users reviewed both games\n- No per-node neighbor cap — full topology preserved\n- Edge weights range from 2 to 420,410 shared reviewers\n\n### steam_all_2005.json (game catalog)\n82,928 games released 2005-2025 with:\n- Title, release year, positive review ratio, total review count\n- Price, developer, rating category\n- Genre indices and tag indices (referencing lookup tables in the file)\n\nPacked as arrays for compact JSON: `[name, year, ratio, reviews, price, ratingIdx, genreIdxs, tagIdxs, developer]`\n\n### steam_force_layout.json (precomputed layout)\nFruchterman-Reingold positions for the top ~9K nodes. Use as warm-start coordinates for force-directed visualization.\n\n## Sources\n\n- **Game metadata**: [FronkonGames Steam Games Dataset](https://huggingface.co/datasets/FronkonGames/steam-games-dataset) (Jan 2026 snapshot, 122K games)\n- **User reviews**: [artermiloff Steam Reviews 2024](https://www.kaggle.com/datasets/artermiloff/steam-games-reviews-2024) (128M reviews across 80K games, one CSV per game, 2012-June 2024)\n\n## How It Was Built\n\n1. Load game metadata from FronkonGames enriched CSV\n2. Scan 30K+ per-game review CSVs, extract user-game pairs\n3. For each user who reviewed 2+ games, generate all game pairs\n4. Count shared reviewers per pair to get edge weights\n5. Filter by minimum shared reviewers\n\nFull pipeline code: [github.com/lukeslp/steam-game-network](https://github.com/lukeslp/steam-game-network)\n\n## Live Visualization\n\n[dr.eamer.dev/datavis/interactive/steam-network/](https://dr.eamer.dev/datavis/interactive/steam-network/)\n\nSix interactive views: scatter plot, chord diagram, force-directed network, streamgraph, heatmap, and treemap. All Canvas-rendered for performance at 80K+ data points.\n\n## License\n\nCC-BY-4.0\n\n## Author\n\nLuke Steuber — [lukesteuber.com](https://lukesteuber.com)",
+  "isPrivate": false,
+  "keywords": ["games", "network-analysis", "steam", "graph-data", "gaming"],
+  "licenses": [{"name": "CC-BY-4.0"}]
+}

steam_analysis.ipynb

@@ -0,0 +1,398 @@
+{
+ "nbformat": 4,
+ "nbformat_minor": 5,
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "name": "python",
+   "version": "3.11.0"
+  }
+ },
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "intro",
+   "metadata": {},
+   "source": [
+    "# Steam Game Network: Catalog Analysis\n",
+    "\n",
+    "Exploratory analysis of 82,928 Steam games (2005\u20132025) from `steam_all_2005.json`.\n",
+    "Uses the packed array format: `[name, year, ratio, reviews, price, ratingIdx, genreIdxs, tagIdxs, developer]`\n",
+    "\n",
+    "**Dataset**: [github.com/lukeslp/steam-network-data](https://github.com/lukeslp/steam-network-data)  \n",
+    "**Live viz**: [dr.eamer.dev/datavis/interactive/steam/](https://dr.eamer.dev/datavis/interactive/steam/)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "imports",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import json\n",
+    "import pandas as pd\n",
+    "import matplotlib.pyplot as plt\n",
+    "import matplotlib.ticker as mticker\n",
+    "import numpy as np\n",
+    "from collections import Counter\n",
+    "\n",
+    "plt.style.use('seaborn-v0_8-whitegrid')\n",
+    "plt.rcParams['figure.figsize'] = (14, 5)\n",
+    "plt.rcParams['font.size'] = 11\n",
+    "STEAM_BLUE = '#1b2838'\n",
+    "STEAM_LIGHT = '#66c0f4'\n",
+    "print('Libraries loaded')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "load-h",
+   "metadata": {},
+   "source": [
+    "## 1. Load Catalog"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "load",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with open('steam_all_2005.json') as f:\n",
+    "    catalog = json.load(f)\n",
+    "\n",
+    "genres = catalog['genres']   # index -> genre name\n",
+    "tags = catalog['tags']       # index -> tag name\n",
+    "ratings = catalog['ratings'] # index -> rating label\n",
+    "games_raw = catalog['games']\n",
+    "\n",
+    "# Unpack array format\n",
+    "# [name, year, ratio, reviews, price, ratingIdx, genreIdxs, tagIdxs, developer]\n",
+    "records = []\n",
+    "for g in games_raw:\n",
+    "    if len(g) < 9:\n",
+    "        continue\n",
+    "    records.append({\n",
+    "        'name': g[0],\n",
+    "        'year': g[1],\n",
+    "        'ratio': g[2],\n",
+    "        'reviews': g[3],\n",
+    "        'price': g[4],\n",
+    "        'rating': ratings[g[5]] if g[5] is not None and g[5] < len(ratings) else 'Unknown',\n",
+    "        'genres': [genres[i] for i in (g[6] or []) if i < len(genres)],\n",
+    "        'tags': [tags[i] for i in (g[7] or []) if i < len(tags)],\n",
+    "        'developer': g[8],\n",
+    "    })\n",
+    "\n",
+    "df = pd.DataFrame(records)\n",
+    "print(f'Games loaded: {len(df):,}')\n",
+    "print(f'Years: {df.year.min()} - {df.year.max()}')\n",
+    "print(f'Genres: {len(genres)}, Tags: {len(tags)}, Ratings: {len(ratings)}')\n",
+    "df.head(3)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "year-h",
+   "metadata": {},
+   "source": [
+    "## 2. Release Year Trend (2005\u20132025)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "year",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "year_counts = df[df['year'].between(2005, 2025)]['year'].value_counts().sort_index()\n",
+    "\n",
+    "fig, ax = plt.subplots(figsize=(14, 5))\n",
+    "ax.bar(year_counts.index, year_counts.values, color=STEAM_LIGHT, edgecolor=STEAM_BLUE, linewidth=0.5)\n",
+    "ax.set_title('Steam Games Released per Year (2005\u20132025)', fontsize=14, fontweight='bold', pad=12)\n",
+    "ax.set_xlabel('Year')\n",
+    "ax.set_ylabel('Games Released')\n",
+    "ax.yaxis.set_major_formatter(mticker.FuncFormatter(lambda x, _: f'{x:,.0f}'))\n",
+    "plt.xticks(year_counts.index, rotation=45)\n",
+    "plt.tight_layout()\n",
+    "plt.show()\n",
+    "print(year_counts.to_string())"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "rating-h",
+   "metadata": {},
+   "source": [
+    "## 3. Rating Distribution"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "rating",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "rating_order = [\n",
+    "    'Overwhelmingly Positive', 'Very Positive', 'Positive', 'Mostly Positive',\n",
+    "    'Mixed', 'Mostly Negative', 'Negative', 'Very Negative', 'Overwhelmingly Negative', 'Unknown'\n",
+    "]\n",
+    "rating_counts = df['rating'].value_counts()\n",
+    "ordered = pd.Series({r: rating_counts.get(r, 0) for r in rating_order if r in rating_counts.index})\n",
+    "\n",
+    "colors_map = {\n",
+    "    'Overwhelmingly Positive': '#1a9850', 'Very Positive': '#52ae32',\n",
+    "    'Positive': '#91cf60', 'Mostly Positive': '#d9ef8b',\n",
+    "    'Mixed': '#fee08b',\n",
+    "    'Mostly Negative': '#fc8d59', 'Negative': '#e34a33',\n",
+    "    'Very Negative': '#b30000', 'Overwhelmingly Negative': '#7f0000', 'Unknown': '#cccccc'\n",
+    "}\n",
+    "colors = [colors_map.get(r, '#999') for r in ordered.index]\n",
+    "\n",
+    "fig, ax = plt.subplots(figsize=(14, 5))\n",
+    "bars = ax.bar(ordered.index, ordered.values, color=colors, edgecolor='white', linewidth=0.5)\n",
+    "ax.set_title('Games by Review Rating', fontsize=14, fontweight='bold', pad=12)\n",
+    "ax.set_ylabel('Game Count')\n",
+    "ax.yaxis.set_major_formatter(mticker.FuncFormatter(lambda x, _: f'{x:,.0f}'))\n",
+    "plt.xticks(rotation=30, ha='right')\n",
+    "for bar, val in zip(bars, ordered.values):\n",
+    "    ax.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 50, f'{val:,}', ha='center', fontsize=8)\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "genre-h",
+   "metadata": {},
+   "source": [
+    "## 4. Top Genres by Game Count and Review Volume"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "genres",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "genre_counts = Counter(g for genres_list in df['genres'] for g in genres_list)\n",
+    "genre_review_vol = {}\n",
+    "for _, row in df.iterrows():\n",
+    "    for g in row['genres']:\n",
+    "        genre_review_vol[g] = genre_review_vol.get(g, 0) + (row['reviews'] or 0)\n",
+    "\n",
+    "top_n = 15\n",
+    "top_genres_count = pd.Series(genre_counts).nlargest(top_n).sort_values()\n",
+    "top_genres_vol = pd.Series(genre_review_vol).nlargest(top_n).sort_values()\n",
+    "\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(16, 7))\n",
+    "top_genres_count.plot.barh(ax=ax1, color=STEAM_LIGHT, edgecolor=STEAM_BLUE, linewidth=0.5)\n",
+    "ax1.set_title('Top Genres by Game Count', fontsize=13, fontweight='bold')\n",
+    "ax1.xaxis.set_major_formatter(mticker.FuncFormatter(lambda x, _: f'{x:,.0f}'))\n",
+    "ax1.set_xlabel('Games')\n",
+    "\n",
+    "top_genres_vol.plot.barh(ax=ax2, color='#FF6B35', edgecolor='white', linewidth=0.5)\n",
+    "ax2.set_title('Top Genres by Total Review Volume', fontsize=13, fontweight='bold')\n",
+    "ax2.xaxis.set_major_formatter(mticker.FuncFormatter(lambda x, _: f'{x/1e6:.0f}M'))\n",
+    "ax2.set_xlabel('Total Reviews')\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "price-h",
+   "metadata": {},
+   "source": [
+    "## 5. Price Distribution"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "price",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_priced = df.dropna(subset=['price'])\n",
+    "free = (df_priced['price'] == 0).sum()\n",
+    "paid = (df_priced['price'] > 0).sum()\n",
+    "\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(14, 5))\n",
+    "\n",
+    "# Free vs paid pie\n",
+    "ax1.pie([free, paid], labels=['Free', 'Paid'], autopct='%1.1f%%', colors=['#4CAF50', STEAM_LIGHT], startangle=90)\n",
+    "ax1.set_title(f'Free vs Paid (n={len(df_priced):,})', fontsize=13, fontweight='bold')\n",
+    "\n",
+    "# Paid price histogram\n",
+    "paid_prices = df_priced[df_priced['price'] > 0]['price'].clip(0, 60)\n",
+    "ax2.hist(paid_prices, bins=40, color=STEAM_LIGHT, edgecolor=STEAM_BLUE, linewidth=0.3)\n",
+    "ax2.set_title('Paid Game Price Distribution', fontsize=13, fontweight='bold')\n",
+    "ax2.set_xlabel('Price (USD)')\n",
+    "ax2.set_ylabel('Count')\n",
+    "ax2.yaxis.set_major_formatter(mticker.FuncFormatter(lambda x, _: f'{x:,.0f}'))\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()\n",
+    "print(f'Free games: {free:,} ({free/len(df_priced)*100:.1f}%)')\n",
+    "print(f'Paid games: {paid:,} ({paid/len(df_priced)*100:.1f}%)')\n",
+    "print(f'Median paid price: ${paid_prices.median():.2f}')\n",
+    "print(f'Mean paid price: ${paid_prices.mean():.2f}')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "top25-h",
+   "metadata": {},
+   "source": [
+    "## 6. Top 25 Games by Review Count"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "top25",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "top25 = df.nlargest(25, 'reviews')[['name', 'year', 'reviews', 'ratio', 'rating', 'price']].copy()\n",
+    "top25['reviews_M'] = (top25['reviews'] / 1e6).round(2)\n",
+    "top25['price_str'] = top25['price'].apply(lambda x: 'Free' if x == 0 else f'${x:.2f}' if pd.notna(x) else 'N/A')\n",
+    "\n",
+    "fig, ax = plt.subplots(figsize=(14, 9))\n",
+    "colors = [('#1a9850' if 'Positive' in r else '#e34a33' if 'Negative' in r else '#fee08b') for r in top25['rating']]\n",
+    "bars = ax.barh(top25['name'][::-1], top25['reviews'][::-1], color=colors[::-1], edgecolor='white', linewidth=0.3)\n",
+    "ax.set_title('Top 25 Steam Games by Total Review Count', fontsize=14, fontweight='bold', pad=12)\n",
+    "ax.set_xlabel('Total Reviews')\n",
+    "ax.xaxis.set_major_formatter(mticker.FuncFormatter(lambda x, _: f'{x/1e6:.1f}M'))\n",
+    "\n",
+    "for bar, (_, row) in zip(bars[::-1], top25.iterrows()):\n",
+    "    ax.text(bar.get_width() + 5000, bar.get_y() + bar.get_height()/2,\n",
+    "            f'{row[\"ratio\"]:.0f}% pos | {row[\"price_str\"]}', va='center', fontsize=8)\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()\n",
+    "\n",
+    "print(top25[['name','year','reviews_M','ratio','price_str']].to_string(index=False))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "genre-evo-h",
+   "metadata": {},
+   "source": [
+    "## 7. Genre Evolution Over Time (Stacked Area)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "genre-evo",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "top_genres = [g for g, _ in Counter(g for genres_list in df['genres'] for g in genres_list).most_common(8)]\n",
+    "df_year = df[df['year'].between(2005, 2025)].copy()\n",
+    "\n",
+    "genre_year = pd.DataFrame(index=range(2005, 2026), columns=top_genres, data=0)\n",
+    "for _, row in df_year.iterrows():\n",
+    "    for g in row['genres']:\n",
+    "        if g in genre_year.columns:\n",
+    "            genre_year.loc[row['year'], g] += 1\n",
+    "\n",
+    "fig, ax = plt.subplots(figsize=(14, 6))\n",
+    "genre_year.plot.area(ax=ax, alpha=0.8, cmap='tab10')\n",
+    "ax.set_title('Genre Composition by Year (Game Count)', fontsize=14, fontweight='bold', pad=12)\n",
+    "ax.set_xlabel('Year')\n",
+    "ax.set_ylabel('Games Released')\n",
+    "ax.yaxis.set_major_formatter(mticker.FuncFormatter(lambda x, _: f'{x:,.0f}'))\n",
+    "ax.legend(loc='upper left', fontsize=9, ncol=2)\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "tags-h",
+   "metadata": {},
+   "source": [
+    "## 8. Tag Co-occurrence Heatmap (Top 20 Tags)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "tag-heatmap",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "tag_counts = Counter(t for tags_list in df['tags'] for t in tags_list)\n",
+    "top20_tags = [t for t, _ in tag_counts.most_common(20)]\n",
+    "\n",
+    "cooccur = np.zeros((20, 20), dtype=int)\n",
+    "for tags_list in df['tags']:\n",
+    "    relevant = [t for t in tags_list if t in top20_tags]\n",
+    "    for i, t1 in enumerate(top20_tags):\n",
+    "        if t1 in relevant:\n",
+    "            for j, t2 in enumerate(top20_tags):\n",
+    "                if t2 in relevant:\n",
+    "                    cooccur[i, j] += 1\n",
+    "\n",
+    "# Normalize diagonal to 1 for readability\n",
+    "diag = np.diag(cooccur)\n",
+    "cooccur_norm = cooccur / diag[:, None]\n",
+    "np.fill_diagonal(cooccur_norm, 1.0)\n",
+    "\n",
+    "fig, ax = plt.subplots(figsize=(12, 10))\n",
+    "im = ax.imshow(cooccur_norm, cmap='YlOrRd', aspect='auto', vmin=0, vmax=0.8)\n",
+    "ax.set_xticks(range(20))\n",
+    "ax.set_yticks(range(20))\n",
+    "ax.set_xticklabels(top20_tags, rotation=45, ha='right', fontsize=8)\n",
+    "ax.set_yticklabels(top20_tags, fontsize=8)\n",
+    "ax.set_title('Tag Co-occurrence (normalized by row tag frequency)', fontsize=13, fontweight='bold', pad=12)\n",
+    "plt.colorbar(im, ax=ax, shrink=0.8, label='Co-occurrence rate')\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "summary-h",
+   "metadata": {},
+   "source": [
+    "## Summary Statistics"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "summary",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "print('=== Steam Catalog Summary ===')\n",
+    "print(f'Total games: {len(df):,}')\n",
+    "print(f'Year range: {df.year.min()} - {df.year.max()}')\n",
+    "print(f'Unique developers: {df.developer.nunique():,}')\n",
+    "print(f'\\nTop rating categories:')\n",
+    "for rating, count in df.rating.value_counts().head(6).items():\n",
+    "    print(f'  {rating}: {count:,} ({count/len(df)*100:.1f}%)')\n",
+    "print(f'\\nReviews:')\n",
+    "print(f'  Total: {df.reviews.sum():,.0f}')\n",
+    "print(f'  Median per game: {df.reviews.median():,.0f}')\n",
+    "print(f'  Games with 0 reviews: {(df.reviews == 0).sum():,}')\n",
+    "print(f'  Games with 100K+ reviews: {(df.reviews >= 100000).sum():,}')"
+   ]
+  }
+ ]
+}