Automated raman-analyzer

The long-term goal of this script is to automate peak-fitting process of Raman Spectrum.
The typical procedure to process raman spectrum would include:

• Import the data
• Smooth the line
• Baseline reduction
• Peak and Shoulder Peak finder
• Peak Fitting

Dependencies

The script is based on numpy,scipy,rampy,matplotlib,lmfit to do data analysis and visualization.

pip install rampy
pip install lmfit

# put the following code in your .bashrc
export PYTHONPATH=$PYTHONPATH:/path/to/raman-analyzer

Usage

Operations on single Raman Spectrum

After running the codes, call:

import analyzer as raman
raman = raman.raman_analyzer(name,min,max,filter,noise_min,noise_max)

• name: your Raman datafile name with full directory (string)

• min,max : the range of wavenumbers you are interested in analyzing

• filter: the filtration threshold for Raman peaks points (the more the filter number is , the less peak points would be stored).

• noise_min,noise_max: the range of wavenumbers you are interested in analyzing the noise level

• raman would be the object in which all the information is stored.

Batch operation of exporting baseline corrected Raman spectra

directory = "raman_spectra_folder"
min = 1000
max = 3600
filter = 8
noise_min = 2000
noise_max = 2200
peaks,signals,fwhm,fwhm_test,name,peak_threshold = raman.raman_batch(directory,min,max,filter,noise_test=True,show=True,noise_min,noise_max)

• noise_test: if True, the script would plot the noise level of each spectrum

• show: if True, the script would plot the baseline corrected Raman spectra

Outputs:

• peaks: the peak positions of each spectrum
• signals: the intensities of peaks of each spectrum
• fwhm: the full width at half maximum of each peak
• name: the name of each spectrum
• peak_threshold: the threshold of peak detection of each spectrum

Calculation of peak ratios in batch

Peak rations of Raman spectra is a useful tool to analyze graphene and other 2D materials.

# "total" item is necessary for the calculation of peak ratios.
peak_param = {"d": [1340,1360,'auto'],
              "g": [1500,1750,9],
              "2d":[2500,2900,9],
              "total":[1000,3600,0.1,2000,2200]}
directory = "./raman_spectra"
peak_data, all_peaks = raman.ratio_calculator(directory,ref ='g',**peak_param)

• directory: the directory of the Raman spectra
• ref: the reference peak for the calculation of peak intensity ratios
• peak_param: the dictionary of peak parameters.
  • "total": it contains the range of the whole spectrum, the filter number, and the range of noise level.
  • All other items: it contains the range of the peak, and the filter number. if the third value in the list is "auto", the script would find the peak using maximum value in the range.

Automate peak fitting on single Raman spectrum

obj_fitting = raman.raman_fitting(name,fit_min,fit_max,kw_fn,\
                                  fit_algo,fit_type,peak_num,\
                                  min,max,filter)

• name,min,max,filter: those are inherited from raman_analyzer class
• fit_min,fit_max: the range of wavenumbers you are interested in fitting
• kw_fn: **kwargs containing parameters for fitting algorithm
• fit_algo: the fitting algorithm you want to use. For details, please refer to lmfit documentation.
• fit_type: the type of fitting peaks. It can be gaussian or lorentzian.
• peak_num: the number of fitting peaks.

Batch operation of exporting peak fitting results

By running the following code, the script would export the fitting parameters of each spectrum in a csv file.

result = raman.raman_fitting_batch(directory,fit_min,fit_max,peak_num,min,max,export,filter,fit_algo,fit_type)

• export: if True, the script would export the fitting peak images

To-do list

• Batch operation of exporting baseline corrected Raman spectra
• Batch operation of exporting peak fitting results
• CLI usage of the script

Author

• Wentong Zhou@Linkedin

Citation

If you use this code, please cite the following paper: