UV-Vis Spectrum Gaussian Fitting Tool

A Python tool for analyzing UV-Vis spectroscopy data by converting wavelengths to wavenumbers, fitting multiple Gaussian peaks, and visualizing the results.

Features

• Automatic unit conversion: Converts wavelength (nm) to wavenumber (cm⁻¹) for fitting
• Multiple Gaussian fitting: Fit any number of Gaussian peaks to your spectrum
• Flexible initial conditions: Specify initial peak positions or use automatic distribution
• Range restriction: Analyze only a specific wavelength range of your spectrum
• Dual visualization: View results in either wavelength (nm) or wavenumber (cm⁻¹) units
• Fit quality metrics: Get R² and RMSE values for fit assessment
• Export capabilities: Save high-resolution plots of your results

Installation

1. Clone or download this repository to your local machine

2. Install Python dependencies:

   pip install -r requirements.txt

   Or install manually:

   pip install numpy pandas matplotlib scipy

Input Data Format

The tool expects a CSV file with two columns (no headers):

1. Column 1: Wavelength in nanometers (nm)
2. Column 2: Absorbance values

Example:

1100.011719,0.007490207
1099.009399,-0.019240458
1098.006226,-0.004369807
...

Usage

Basic Syntax

python uvvis_gaussian_fit.py <csv_file> <n_gaussians> [options]

Parameters

• csv_file: Path to your CSV data file
• n_gaussians: Number of Gaussian peaks to fit (integer)

Options

• --initial_peaks PEAK1 PEAK2 ...: Specify initial peak positions in wavelength (nm)
• --initial_peaks_cm PEAK1 PEAK2 ...: Specify initial peak positions in wavenumber (cm⁻¹)
• --range_nm MIN MAX: Restrict analysis to wavelength range (nm)
• --display_unit {nm,cm}: Choose display unit for final plot (default: nm)
• --save FILENAME: Save plot to file (PNG, PDF, etc.)

Examples

1. Basic Usage - Automatic Peak Distribution

python uvvis_gaussian_fit.py spectrum.csv 3

Fits 3 Gaussians with automatically distributed initial peak positions.

2. Specify Initial Peak Positions (Wavelength)

python uvvis_gaussian_fit.py spectrum.csv 2 --initial_peaks 400 500

Fits 2 Gaussians with initial centers at 400 and 500 nm.

2b. Specify Initial Peak Positions (Wavenumber)

python uvvis_gaussian_fit.py spectrum.csv 2 --initial_peaks_cm 25000 20000

Fits 2 Gaussians with initial centers at 25,000 and 20,000 cm⁻¹.

3. Analyze Specific Wavelength Range

python uvvis_gaussian_fit.py spectrum.csv 3 --range_nm 300 700

Only analyzes data between 300-700 nm, fitting 3 Gaussians.

4. Display in Wavenumber Units

python uvvis_gaussian_fit.py spectrum.csv 2 --display_unit cm

Shows the final plot with wavenumber (cm⁻¹) on the x-axis.

5. Save High-Resolution Plot

python uvvis_gaussian_fit.py spectrum.csv 3 --save my_spectrum_fit.png

Saves the result plot as a high-resolution PNG file.

6. Complete Example

python uvvis_gaussian_fit.py spectrum.csv 3 \
  --initial_peaks 360 450 625 \
  --range_nm 350 650 \
  --display_unit nm \
  --save spectrum_analysis.pdf

Output Information

The tool provides detailed output including:

Fit Quality Metrics

• R²: Coefficient of determination (closer to 1.0 = better fit)
• RMSE: Root mean square error (smaller = better fit)

Gaussian Parameters

For each fitted Gaussian:

• Center: Peak position in both cm⁻¹ and nm
• Amplitude: Peak height (absorbance units)
• Width: Peak width in both cm⁻¹ and nm

Visual Output

• Original data points
• Sum of all Gaussians (fitted curve)
• Individual Gaussian components
• Fit quality metrics in plot title

Tips for Best Results

Choosing Number of Gaussians

• Start with fewer Gaussians (2-3) and increase if needed
• Too many Gaussians may lead to overfitting
• Use physical/chemical knowledge to guide the choice

Initial Peak Positions

• Examine your spectrum visually to identify approximate peak positions
• Easy method: Use --initial_peaks with wavelengths in nm (e.g., 400 500 600)
• Advanced method: Use --initial_peaks_cm with wavenumbers in cm⁻¹
• Conversion: wavenumber = 10,000,000 / wavelength_nm (e.g., 500 nm → 20,000 cm⁻¹)

Wavelength Range Selection

• Exclude noisy regions at spectrum edges
• Focus on the region of interest for your analysis
• Ensure sufficient data points for reliable fitting

Troubleshooting Poor Fits

• Try different initial peak positions
• Adjust the number of Gaussians
• Check data quality and remove outliers if necessary
• Consider if Gaussian peaks are appropriate for your system

Example Workflow

1. Examine your data:

   # Quick visual inspection
   python -c "import pandas as pd; import matplotlib.pyplot as plt; 
   data=pd.read_csv('spectrum.csv', header=None); 
   plt.plot(data[0], data[1]); plt.show()"

2. Start with basic fit:

   python uvvis_gaussian_fit.py spectrum.csv 2

3. Refine based on results:

   python uvvis_gaussian_fit.py spectrum.csv 3 --initial_peaks 25000 20000 15000

4. Final analysis with range restriction:

   python uvvis_gaussian_fit.py spectrum.csv 3 \
     --initial_peaks 25000 20000 15000 \
     --range_nm 400 800 \
     --save final_analysis.png

Understanding the Mathematics

Unit Conversion

• Wavelength to Wavenumber: ν̃ = 10⁷/λ (cm⁻¹ = 10⁷ nm·cm⁻¹ / nm)
• Wavenumber to Wavelength: λ = 10⁷/ν̃ (nm = 10⁷ nm·cm⁻¹ / cm⁻¹)

Gaussian Function

Each Gaussian peak is defined as:

G(ν̃) = A × exp(-(ν̃ - ν̃₀)²/(2σ²))

Where:

• A = amplitude (peak height)
• ν̃₀ = center position (cm⁻¹)
• σ = width parameter (cm⁻¹)

Multi-Gaussian Fit

The total spectrum is the sum of individual Gaussians:

S(ν̃) = Σ Aᵢ × exp(-(ν̃ - ν̃₀ᵢ)²/(2σᵢ²))

File Structure

uvvis_convertfit/
├── uvvis_gaussian_fit.py    # Main analysis script
├── requirements.txt         # Python dependencies
├── examples.py             # Usage examples
├── README.md              # This documentation
└── spectrum.csv            # Example data file

Support and Contributing

If you encounter issues or have suggestions for improvements:

1. Check that your CSV file format matches the expected structure
2. Verify all dependencies are installed correctly
3. Try simpler cases first (fewer Gaussians, no range restrictions)

License

This tool is provided as-is for scientific and educational use.