select the max element in an increasing (and restarting over) list

I have this list:

mylist = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 
          1, 1, 1, 2, 1, 1, 1, 1, 2, 1, 1, 2, 1, 2, 3]

The length of the list is : 38.

I want a new list with length: 32.

My new list will be:

new_list = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 
            1, 1, 2, 1, 1, 1, 2, 1, 2, 3]

So, the numbers represent different counts. The first 19 elements of the mylist are all 1. The 20 and 21 elements are 1 and 2. This means I had 2 counts.

I want to get rid of the 1 and keep only the 2 because I have 2 counts. The same goes to the next 1 and 2 elements and to the last 1, 2, 3 elements. There, I have a max count of 3.

If I try something like this:

new_list = []
for idx in range(len(mylist)-1):
    if mylist[idx] == 1:
        if mylist[idx + 1] > mylist[idx]:
            new_list.append(mylist[idx + 1])
        else:
            new_list.append(mylist[idx])

because I am only checking the current and next elements, I miss the last 3.

>Solution :

Providing the first element in mylist is less than 2 then:

mylist = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 
          1, 1, 1, 2, 1, 1, 1, 1, 2, 1, 1, 2, 1, 2, 3]
newlist = []

for x in mylist:
    if x > 1:
        newlist[-1] = x
    else:
        newlist.append(x)

print(newlist)