scripts to deal with data acquired at different (or non-uniform) sample rates.

discreet_time.py

import numpy as np
import matplotlib.pyplot as plt
plt.ion()
import scipy.interpolate

def interpolate_to_common_samplerate(Y, X=None, t=None):
    if t is None:
        if X is None:
            t = np.arange(len(Y[0]),dtype='float')
        else:
            t = X[0]
    interpolatedY = np.zeros((len(t),len(Y)),dtype='float')        
    for i, y in enumerate(Y):
        if X is None:
            x = np.linspace(t[0],t[-1],len(y),endpoint=True)
        else:
            x = X[i]
        sortedInds = np.argsort(x)
        x = x[sortedInds]
        y = y[sortedInds]
        f = scipy.interpolate.interp1d(x, y, bounds_error=False)
        interpolatedY[:,i] = f(t)
        
    return interpolatedY, t
        
    
T=1000
n1=100
t1=np.linspace(0,T,n1,endpoint=True)+np.random.rand(n1)*10
t1[-1]=T
y1=np.random.rand(n1)+t1/20.
plt.plot(t1,y1,'.k')

n2=90
t2=np.linspace(0,T,n2,endpoint=True)
y2=np.random.rand(n2)+t2/20.+1
plt.plot(t2,y2,'.b')

n3=80
t3=np.linspace(0,T,n3,endpoint=True)
y3=np.random.rand(n3)+t3/20.+2
plt.plot(t3,y3,'.g')


#ys, t = interpolate_to_common_samplerate([y1, y2, y3])
#plt.plot(t,ys)

#ys, t = interpolate_to_common_samplerate([y1, y2, y3],t=np.arange(0,1000))
#plt.plot(t,ys)

ys, t = interpolate_to_common_samplerate([y1, y2, y3],[t1, t2, t3],t=np.linspace(0,T,10000,endpoint=True))
plt.plot(t,ys)