Accessing the position [i-1] of an array in python - Filter

Question

I'm trying to set up a filter in Python, and I am stuck in a basic stuff.

This is the function that I have to implement:

y[n] = (1 − λ) · x[n] + λ · y[n − 1]

I'm reading an m file and applying the filter.

Here's the code:

import sys
import numpy as np
import matplotlib.pyplot as plt

y = np.loadtxt('acs712_192us.m')
size = len(y)

x = np.arange(0, size)

out = []

lamb = 0.9 

for i in range(0, len(y)):
    out.append(((1-lamb) * y[i]) + (lamb * out[i - 1]))

plt.plot(x, y)
plt.plot(x, out)
plt.show()

When I run this program, get the following error:

File "main_LI.py", line 15, in out.append(((1-lamb) * y[i]) + (lamb * out[i - 1])) IndexError: list index out of range

I know i can't access out[-1] position (in the first loop), maybe is this the problem?

Anyone have any tips?

Thank you guys!

Your recurrence relation doesn't seem to have a starting point. — user2357112, May 22 '18 at 16:25
I doubt there should be an initial value in `out` to start with. — Austin, May 22 '18 at 16:27
The file is a .txt with 512 values (0 - 511). I can get the y[i] — Matheus Bernardi, May 22 '18 at 16:28
out[0] = (1 - 0.9) * y[0] + 0.9*out[-1] I need ignore this first loop (i guess), cause this access out[-1] — Matheus Bernardi, May 22 '18 at 16:32

score 1 · Accepted Answer · answered May 22 '18 at 16:32

First, you can use negative indexes in Python. The problem is that you are providing an index that does not exist.

out initially is an empty list, so the problem is that you are trying to access an index that does not exist in the first iteration:

import sys
import numpy as np
import matplotlib.pyplot as plt

y = np.loadtxt('acs712_192us.m')
size = len(y)

x = np.arange(0, size)

out = []

lamb = 0.9 

for i in range(len(y)):
    if i == 0:
       out.append(value) # value is what you want out[0] to be
    else:
        out.append(((1-lamb) * y[i]) + (lamb * out[i - 1]))

plt.plot(x, y)
plt.plot(x, out)
plt.show()

It works! Thanks for the help, sorry if was a stupid question. — Matheus Bernardi, May 22 '18 at 16:36

score 0 · Answer 2 · answered May 22 '18 at 16:35

0

You can use list comprehension to do this pretty easily.

l = 0.9
x = np.arange(0, 10)
filtered = [x[i] if i == 0 else (1-l)*x[i] + l*x[i-1] for i in range(0,len(x))]

>>> print(filtered)
[0, 0.09999999999999998, 1.1, 2.1, 3.1, 4.1, 5.1, 6.1000000000000005, 7.1, 8.1]

is this the output you expect?

answered May 22 '18 at 16:35

Grant Williams

1,469
1
12
24

That would help too! Thanks – Matheus Bernardi May 22 '18 at 16:39

Accessing the position [i-1] of an array in python - Filter

2 Answers2