n8n-nodes-sklearn

Custom n8n nodes for integrating scikit-learn machine learning algorithms into your n8n workflows.

Features

This package provides n8n nodes for popular scikit-learn functionality:

• Sklearn Linear Regression: Train and predict using linear regression models
• Sklearn Standard Scaler: Standardize features by removing the mean and scaling to unit variance

Installation

Prerequisites

1. Python 3.7+ with scikit-learn installed:

   pip install scikit-learn numpy

2. n8n installed (version 0.190.0 or higher)

Installing the Package

Option 1: Install from npm (after publishing)

npm install n8n-nodes-sklearn

Option 2: Local Development Installation

1. Clone or download this repository

2. Navigate to the package directory:

   cd n8n-nodes-sklearn

3. Install dependencies:

   npm install

4. Build the package:

   npm run build

5. Link the package to your n8n installation:

   npm link

6. In your n8n custom nodes directory (usually ~/.n8n/custom):

   npm link n8n-nodes-sklearn

7. Restart n8n to load the new nodes

Usage

Sklearn Linear Regression

Train Operation

Train a linear regression model using your input data.

Input Data Format:

[
  {
    "feature1": 1.0,
    "feature2": 2.0,
    "feature3": 3.0,
    "target": 10.5
  },
  {
    "feature1": 2.0,
    "feature2": 3.0,
    "feature3": 4.0,
    "target": 15.2
  }
]

Parameters:

• Feature Columns: Comma-separated list of feature column names (e.g., feature1,feature2,feature3)
• Target Column: Name of the target column (e.g., target)
• Fit Intercept: Whether to calculate the intercept (default: true)
• Python Path: Path to Python executable (default: python3)

Output:

{
  "model": "{...}",
  "coefficients": [1.2, 3.4, 2.1],
  "intercept": 0.5,
  "r2_score": 0.95,
  "feature_columns": ["feature1", "feature2", "feature3"],
  "training_samples": 100
}

Predict Operation

Make predictions using a trained model.

Parameters:

• Model Data: JSON string containing the trained model (from train operation)
• Feature Columns: Comma-separated list of feature columns (must match training)
• Python Path: Path to Python executable

Output: The original input data with an added prediction field.

Sklearn Standard Scaler

Fit Transform Operation

Fit the scaler to your data and transform it in one step.

Input Data Format:

[
  {
    "age": 25,
    "income": 50000,
    "score": 85
  },
  {
    "age": 35,
    "income": 75000,
    "score": 92
  }
]

Parameters:

• Feature Columns: Comma-separated list of columns to scale (e.g., age,income,score)
• With Mean: Whether to center the data (default: true)
• With Std: Whether to scale to unit variance (default: true)
• Output Prefix: Prefix for scaled columns (default: scaled_)
• Python Path: Path to Python executable

Output: Original data with scaled features added:

{
  "age": 25,
  "income": 50000,
  "score": 85,
  "scaled_age": -0.707,
  "scaled_income": -0.707,
  "scaled_score": -0.707,
  "scaler": "{...}",
  "scaler_info": {
    "mean": [30, 62500, 88.5],
    "scale": [7.071, 17677.67, 4.95]
  }
}

Fit Operation

Fit the scaler and save parameters for later use.

Output:

{
  "scaler": "{...}",
  "mean": [30, 62500, 88.5],
  "scale": [7.071, 17677.67, 4.95],
  "variance": [50, 312500000, 24.5],
  "feature_columns": ["age", "income", "score"],
  "fitted_samples": 100
}

Transform Operation

Transform data using a previously fitted scaler.

Parameters:

• Scaler Data: JSON string from fit operation
• Feature Columns: Comma-separated list of columns (must match fitted columns)
• Output Prefix: Prefix for scaled columns

Example Workflows

Example 1: Simple Linear Regression Pipeline

1. Read CSV Node: Load your training data
2. Sklearn Linear Regression (Train): Train the model
3. Set Node: Store the model in a variable
4. Read CSV Node: Load new data for prediction
5. Sklearn Linear Regression (Predict): Make predictions

Example 2: Preprocessing Pipeline

1. HTTP Request Node: Fetch data from API
2. Sklearn Standard Scaler (Fit Transform): Normalize features
3. Sklearn Linear Regression (Train): Train on normalized data
4. Code Node: Evaluate model performance

Configuration

Python Path

By default, nodes use python3 command. If you need to specify a different Python executable:

1. Set the Python Path parameter in each node
2. Or set an environment variable before starting n8n:

   export PYTHON_PATH=/path/to/python3
   n8n start

Troubleshooting

Error: Python script failed

• Ensure Python 3.7+ is installed
• Verify scikit-learn is installed: python3 -c "import sklearn; print(sklearn.__version__)"
• Check Python path in node parameters

Error: Feature column not found

• Verify column names match your input data exactly (case-sensitive)
• Check for extra spaces in column names

Memory issues with large datasets

• Consider processing data in batches
• Use n8n's batch processing features
• Increase Node.js memory limit: NODE_OPTIONS=--max-old-space-size=4096 n8n start

Development

Building

npm run build

Linting

npm run lint
npm run lintfix  # Auto-fix issues

Adding New Nodes

1. Create a new directory in nodes/
2. Create YourNode.node.ts implementing INodeType
3. Add the node to package.json under n8n.nodes
4. Run npm run build

Roadmap

Future nodes planned:

• Logistic Regression
• Decision Trees / Random Forest
• K-Means Clustering
• Principal Component Analysis (PCA)
• Support Vector Machines (SVM)
• Model evaluation metrics
• Cross-validation utilities

Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

License

MIT

Support

For issues and questions:

• GitHub Issues: https://github.com/arturovaine/n8n-nodes-sklearn/issues
• n8n Community Forum: https://community.n8n.io/

Acknowledgments

• Built on top of the excellent n8n workflow automation platform
• Uses scikit-learn for machine learning functionality