polypython

Python library for the spectroscopic identification of plastic particulates in Raman spectral images. Specifically, for Raman images collected using Renishaw inVia Raman microscope's StreamLine^TM scanning mode with Raman shifts centred at 1300 cm^-1 and ranging from 924.6 to 1668.0 cm^-1. Polypython will identify ten plastics i.e. Acrylonitrile butadiene styrene, Polyamide, Polycarbonate, Polyethylene terephathalate, Perfluoroalkoxy alkane, Polypropylene, Polystyrene, Polyurethane, Polymethylmethacrylate, and Polyvinyl chloride, two pigements i.e. Cibacron Brilliant Red 3B-A, and Copper phthalocyanine, and one semi-synthetic material i.e. modified cellulose. This python library consists of four spectral normalisation equations, a x and y pixel dimension estimator and a spectral image analysis function. Load the polypython library using the following:

* pip install polypython
import polypython as pp

Spectral normalisation

The four spectral normalisation equations include max, area, interval, and standard normal variate (SNV). All spectral images should be loaded in .txt format with rows representing Raman spectrums and columns being the Raman shifts (cm^-1). The four normalisation codes can be called using the following:

data = pp.norm(data, mode='snv')
data = pp.norm(data, mode='max')
data = pp.norm(data, mode='area')
data = pp.norm(data, mode='01')

xy pixel estimation

An x and y pixel number can be estimated from the spectrum number using prime factorisation by running the following code:

#data shape: [Raman spectrum number, Raman shifts (cm<sup>-1</sup>)
xy = pp.xy(data)

File_locator

To save plastic analysed spectral images in an appropriate location, the user can define where to save the spectral images using the following:

#data shape: [Raman spectrum number, Raman shifts (cm<sup>-1</sup>)
location = pp.file_locator()
#which will print: where would you like to save your data and graphs? i.e. "/home/Desktop/DataFolder" 
/home/Desktop/DataFolder/ #an example of the desired file path

Spectral image visualisation

Visualisation of the unanalysed spectral image can be conducted by calling the following:

#data shape: [Raman spectrum number, Raman shifts (cm<sup>-1</sup>)
#xy: xy image pixel dimension
#location: file location saver
pp.polypython.vis(data, xy, mode='mean')

Polyrsi analysis

The polyrsi function can be used to identify the presence of plastic related spectra in a Raman spectral image using the following:

#data shape: [Raman spectrum number, Raman shifts (cm<sup>-1</sup>)
#xy: xy image pixel dimension
#location: file location saver
pp.polyrsi(data,xy,location)

If you decide to use Polypython to analyse your Raman spectral images, please reference this academic journal: https://doi.org/10.1021/acs.analchem.9b05445.