apodise.py.save

# -*- coding: utf-8 -*-
"""
Created on Wed Sep 12 17:34:35 2018

@author: Peter
"""
from scipy import interpolate, fftpack, signal
import numpy as np
# Function to Fourier filter spectra

class apodiser():
        
    def fftspec(self,spectrum):
    
        fft = fftpack.fft(spectrum[:,1])
        filtered_fft = fft.copy()
        filtered_fft[100:]=0    # Can play around with this value (100) to filter more or less
        filtered_spectrum = fftpack.ifft(filtered_fft)
        spectrum[:,1] = filtered_spectrum
        
        return spectrum
    
    # Function to fit a spline to a spectrum and return a spectrum normalised
    # about zero
    def splinespec(self,spectrum):
    
        spline_fit = interpolate.UnivariateSpline(spectrum[:,0],spectrum[:,1])
        normalised_spectrum = (spectrum[:,1]/spline_fit(spectrum[:,0]))-1
    
        spectrum[:,1] =  normalised_spectrum
        return spectrum
    
    def mfilt (self,spectrum):
        filter = np.ones(len(spectrum))
        length =len(spectrum)
        for i in range(0,int(length*0.1)):
            filter[i] = i/(length*0.1)
        for i in range(int(length*0.90),length):
            filter[i] =  (float(length-i)/((length-(length*0.90))+1))
            
        return filter    
    
    def hanwin(self,spectrum):
        endlength = int(0.05*len(spectrum))
        cosine = signal.hann(endlength*2)
        hann = np.on(len(spectrum))
        hann[0:endlength] = cosine[0:endlength]
        hann[-endlength:] = cosine[endlength:endlength*2]
        hann = spectrum * hann
        return hann
    
    def apodise(self,  spectrum):
        spectrum = self.splinespec(spectrum)
        spectrum = self.fftspec(spectrum)
        spectrum = self.mfilt(spectrum)
        spectrum = self.hanwin(spectrum)
        return spectrum