Installation

Installation Guide

Prerequisites

System Requirements

• Operating System: Windows, macOS, or Linux
• Python: Version 2.7 or 3.6+ (code uses Python 2 pickle format)
• Webcam: Required for emotion detection
• Internet: Required for Algorithmia API calls
• Browser: Modern browser with WebRTC support (Chrome, Firefox, Safari, Edge)

Required Accounts

• Algorithmia Account: Free tier available at algorithmia.com

---

Step-by-Step Installation

1. Clone or Download Repository

# If using git
git clone <repository-url>
cd muze

# Or download and extract ZIP file

2. Install Python Dependencies

Using pip

pip install Flask==0.12.2
pip install hsaudiotag==1.1.1
pip install Jinja2==2.10
pip install matplotlib==2.1.2
pip install numpy==1.14.0
pip install Pillow==5.0.0
pip install simplejson==3.13.2
pip install six==1.11.0
pip install urllib3==1.22
pip install Werkzeug==0.14.1
pip install Algorithmia

Using requirements.txt (if available)

pip install -r requirements.txt

Create requirements.txt

If not present, create requirements.txt:

Flask==0.12.2
hsaudiotag==1.1.1
Jinja2==2.10
matplotlib==2.1.2
numpy==1.14.0
Pillow==5.0.0
simplejson==3.13.2
six==1.11.0
urllib3==1.22
Werkzeug==0.14.1
Algorithmia>=1.0.0

3. Get Algorithmia API Key

1. Visit algorithmia.com
2. Sign up for a free account
3. Navigate to your profile
4. Click on "Credentials" section
5. Copy your API key (starts with "sim...")

4. Configure API Key

Open algorithmia.py and replace the API key on line 17:

# Before
client = Algorithmia.client('api-key')

# After
client = Algorithmia.client('simYOUR_ACTUAL_API_KEY_HERE')

⚠️ Security Note: Never commit your API key to version control!

Better approach (recommended):

Create a .env file:

ALGORITHMIA_API_KEY=simYOUR_ACTUAL_API_KEY_HERE

Update algorithmia.py:

import os
from dotenv import load_dotenv

load_dotenv()
client = Algorithmia.client(os.getenv('ALGORITHMIA_API_KEY'))

Install python-dotenv:

pip install python-dotenv

5. Verify Directory Structure

Ensure these directories exist:

muze/
├── app.py
├── algorithmia.py
├── test.txt
├── snapshots/          # Create if missing
├── static/
│   ├── music/          # Should contain 001.mp3 - 903.mp3
│   └── jquery-1.7.1.min.js
└── templates/
    ├── base.html
    └── musi.html

Create missing directories:

mkdir -p snapshots
mkdir -p static/music

6. Verify Music Files

Check that music files are present:

# Windows
dir static\music\*.mp3

# macOS/Linux
ls static/music/*.mp3

You should see files: 001.mp3, 002.mp3, ..., 903.mp3

If music files are missing: You'll need to obtain them from the MIREX 2007 dataset or provide your own music library.

---

Running the Application

Start the Server

python app.py

Expected output:

 * Running on http://127.0.0.1:5000/ (Press CTRL+C to quit)
 * Restarting with stat
 * Debugger is active!
 * Debugger PIN: XXX-XXX-XXX

Access the Application

1. Open your web browser
2. Navigate to: http://localhost:5000
3. Allow webcam access when prompted
4. Wait 5 seconds for initial snapshot
5. Enjoy your personalized playlist!

---

Troubleshooting Installation

Python Version Issues

Problem: SyntaxError or compatibility issues

Solution: Check Python version

python --version

If using Python 3, ensure all code is compatible or use Python 2.7.

Pickle Loading Error

Problem: UnicodeDecodeError when loading test.txt

Solution: Ensure encoding parameter is used

pickle.load(fp, encoding='latin1')

Module Not Found

Problem: ModuleNotFoundError: No module named 'flask'

Solution: Install missing module

pip install flask

Algorithmia Import Error

Problem: ImportError: No module named Algorithmia

Solution: Install Algorithmia SDK

pip install Algorithmia

Port Already in Use

Problem: OSError: [Errno 48] Address already in use

Solution: Change port in app.py

if __name__ == '__main__':
    app.run(debug=True, port=5001)  # Use different port

Or kill existing process:

# Find process using port 5000
lsof -i :5000

# Kill process
kill -9 <PID>

Webcam Not Working

Problem: "getUserMedia not supported" or permission denied

Solutions:

1. Use HTTPS (required for some browsers)
2. Check browser permissions
3. Ensure webcam is not in use by another app
4. Try different browser

API Key Invalid

Problem: AlgorithmException: authorization required

Solution:

1. Verify API key is correct
2. Check Algorithmia account is active
3. Ensure API key has proper permissions

Missing Music Files

Problem: Audio player shows error or no sound

Solution:

1. Verify MP3 files exist in static/music/
2. Check file naming (001.mp3, not 1.mp3)
3. Ensure files are valid MP3 format

---

Verification Tests

Test 1: Python Environment

python -c "import flask, Algorithmia, PIL, numpy, matplotlib; print('All modules imported successfully')"

Test 2: File Permissions

# Check read permissions
python -c "import pickle; pickle.load(open('test.txt', 'rb'), encoding='latin1'); print('Song database loaded')"

# Check write permissions
python -c "open('snapshots/test.txt', 'w').write('test'); print('Snapshot directory writable')"

Test 3: API Connection

import Algorithmia

client = Algorithmia.client('YOUR_API_KEY')
algo = client.algo('deeplearning/EmotionRecognitionCNNMBP/1.0.1')
print("API connection successful")

Test 4: Server Start

python app.py &
curl http://localhost:5000
# Should return HTML content

---

Docker Installation (Alternative)

Create Dockerfile

FROM python:3.8-slim

WORKDIR /app

COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt

COPY . .

EXPOSE 5000

ENV FLASK_APP=app.py
ENV FLASK_ENV=development

CMD ["python", "app.py"]

Build and Run

# Build image
docker build -t muze .

# Run container
docker run -p 5000:5000 -e ALGORITHMIA_API_KEY=your_key muze

---

Virtual Environment (Recommended)

Create Virtual Environment

# Python 3
python -m venv venv

# Python 2
virtualenv venv

Activate Virtual Environment

# Windows
venv\Scripts\activate

# macOS/Linux
source venv/bin/activate

Install Dependencies

pip install -r requirements.txt

Deactivate

deactivate

---

Production Deployment

Using Gunicorn (Linux/macOS)

# Install Gunicorn
pip install gunicorn

# Run with 4 workers
gunicorn -w 4 -b 0.0.0.0:5000 app:app

Using Waitress (Windows)

# Install Waitress
pip install waitress

# Run server
waitress-serve --port=5000 app:app

Environment Variables

Create .env file for production:

FLASK_ENV=production
FLASK_DEBUG=False
SECRET_KEY=your_secure_random_secret_key
ALGORITHMIA_API_KEY=your_api_key

Security Checklist

• Change Flask secret key from "bacon"
• Move API key to environment variable
• Use HTTPS in production
• Configure CORS properly
• Add rate limiting
• Implement input validation
• Set up logging
• Configure firewall rules

---

Next Steps

After successful installation:

1. Read the Quick Start Guide
2. Review Configuration options
3. Explore API Reference
4. Check Troubleshooting for common issues

---

Uninstallation

Remove Virtual Environment

# Deactivate if active
deactivate

# Remove directory
rm -rf venv

Remove Application

# Delete application directory
rm -rf muze

Uninstall Python Packages

pip uninstall Flask hsaudiotag Jinja2 matplotlib numpy Pillow simplejson six urllib3 Werkzeug Algorithmia