[FIXED] How to find a similar sequence of numbers in a list and return the index offset

Issue

Given 2 lists that are similar but not identical, how can I find the offset of where one list will match the other. They pixel values of a single column of overlapping black and white images so their values will be close but not exact. I have grouped the numbers into 3 "bins" (0,1,2) depending on their value. I am working in Python. In the lists below, list A starts 44 positions (pixels) before list B. The numbers after index 44 are "close" but not exact. The expected output would be 44.

Lists

A [1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2]

B [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

Initial Code

import matplotlib.pyplot as plt
import numpy as np

A = [1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2]
B = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]


def plot_this(aa, bb, idx, rating ):
    xs = [x for x in range(len(aa))]
    fig, ax = plt.subplots(figsize=(8, 4), layout='constrained')
    plt.plot(xs, aa, label ="A")
    plt.plot(xs, bb, label ="B")
    title = str(idx) + " = " + str(rating)
    ax.set_title(title)  # Add a title to the axes.
    ax.legend() 
    plt.show()



x = np.linspace(0, 2 * np.pi, 200)
y = np.sin(x)

fig, ax = plt.subplots()
ax.plot(x, y)
plt.show()
A1 = A.copy()
B1 = B.copy()
score = 0
bad = 0
score_dict = {}
for i in range(len(A1)):
    for j in range(len(A1)):
        if A1[j] == B1[j]:
            score += 1
        else:
            bad += 1
    if bad >= score:
        score_dict[i] = 0
    else:
        rating = (score-bad)/len(A)
        score_dict[i] = rating
        if rating > 0.51:
            plot_this(A1, B1, i, rating)
    score = 0
    bad = 0
    A1.pop(0)
    B1.pop(-1)
for kk, vv in score_dict.items():
    print(f"index {kk}  Score {vv}")

UPDATED CODE

import matplotlib.pyplot as plt
import numpy as np

A = [1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2]
B = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
def plot_this(aa, bb, idx, rating ):
    xs = [x for x in range(len(aa))]
    fig, ax = plt.subplots(figsize=(8, 4), layout='constrained')
    plt.plot(xs, aa, label ="A")
    plt.plot(xs, bb, label ="B")
    title = str(idx) + " = " + str(rating)
    ax.set_title(title)  # Add a title to the axes.
    ax.legend() 
    plt.show()

A1 = A.copy()
B1 = B.copy()
score_dict = {}
for i in range(int(2*len(A1)/3)): #Only go 2/3 way
    coeff = np.corrcoef(A1, B1)
    ranking = coeff[0,1]
    score_dict[i] = coeff[0,1]
    if ranking > 0.9:
            plot_this(A1, B1, i, ranking)
    A1.pop(0) #pop off both ends instead of using numpy.roll()
    B1.pop(-1) #Use numpy.roll() and go all the way around if you don't know which way to go.
for kk, vv in score_dict.items():
    print(f"index {kk}  Score {vv}")
high_score = max(score_dict, key=score_dict.get)
print(high_score)

Final Plot

Solution

Is it okay to use numpy? Then I would recommend having a look at numpy.roll() for shifting the first array and numpy.corrcoef() for computing the correlation between the arrays. If your correlation coefficient is close to 1, you have a good agreement between the two arrays.

Answered By - Cornelia