How can I make a faster search algorithm

[paul18fr]

Hi

Interesting topic (briefly read I confess); let me doing some remarks:
- if you're dealing with numbers (0 and 1 typically), then you can use numpy and vectorization to speed-up your code
- it can be combined with "slicing" to get and compare arrays/vectors
- have a look to numpy.array_equal to compare 2 arrays/vectors (never tested so far but I can imagine the size of the vector is not an issue),

If you cannot avoid loops, have a look to Numba library (http://numba.pydata.org)

Finally, interesting benchmarking has been done by the Nasa to compare C, Python, Julia languages among others, and C is much faster than Python
(type "Basic Comparison of Python, Julia, Matlab, IDL and Java (2018 Edition)" + Nasa on Google); it's also interesting to note than vectorization is very powerfull

Paul

---

n = 1000000;
vect1 = np.random.randint(2,size = n);
vect2 = np.copy(vect1);
identical = np.array_equiv(vect1,vect2);

## now I change 1 value to 2
chang = np.random.randint(n, size = 1);
vect2[chang] = 2;
identical2 = np.array_equiv(vect1,vect2);

extract_part = vect2[100 : 350];

[python]PUT CODE HERE