Reputation: 63
Given a list of integers [0,0,0,1,1,0,2,2,0,1,0,0,2,1,2,2,2,2,1,...]
, I need to calculate the longest subsequence of elements of which x% of the elements are n, i.e. there is a tolerance whereby a subsequence which contains less than 1 - x% values other than n is still counted as an unbroken subsequence of n's.
I have used the following one-liner to get the longest subsequence where all values are n, but I don't know where to go from here:
longest_subsequence_0 = max((len(l) for n, l in itertools.groupby(list) if n == 0))
If anyone could steer me in the right direction it would be much appreciated :D
Upvotes: 2
Views: 271
Reputation: 92
Just a sliding window changing size and computing the %. If above threshold, record the window size if this one is larger than the previous stored one.
def longest_sub(list, n, threshold):
largest_window = 0
for i in range(len(list)+1): # from i
for j in range(i+1,len(list)+1): # to j
window_len = len(list[i: j]) # store window size
if window_len > largest_window: # if inspected window > largest found yet
if list[i:j].count(n)/window_len*100 > threshold: # if percentage above threshold
largest_window = window_len # new largest_window
return largest_window
longest_sub([0,0,0,1,1,0,2,2,0,1,0,0,2,1,2,2,2,2,1], 0, 30) # 9
longest_sub([0,0,0,1,1,0,2,2,0,1,0,0,2,1,2,2,2,2,1], 0, 80) # 3
Upvotes: 1
Reputation: 1375
I have this
import numpy as np
MAX = 10
a = 5
LEN = 100
threshold = 0.3
tolerance = 0.05
x = list(np.random.choice(list(range(MAX)), size=LEN))
window = LEN
largest = 0
while window > 1 and largest == 0:
for i in range(0, LEN-window+1):
ratio_of_a = x[i:i+window].count(a)/window
if ratio_of_a > threshold - tolerance:
largest = window
window -= 1
print(f"Largest sequence found {largest}")
Pretty brute force, i used a
instead of n
also sorry for the confusion :)
Upvotes: 1