I have data that looks like this:
It's standard financial price data (open, high, low, close).
In addition, I run some calculations. 'major_check' occasionally returns 1 or 2 (which 'minor_check' will then also return). 'minor_check' also returns 1 or 2, but more frequently. the rest is filled with 0 or NaN.
I'd like to test for specific patterns:
- Whenever there is a 2 in 'major_check', I want to see if I can find a 21212 pattern in 'minor_check', with 21 preceding the central 2 and 12 following it.
- If there is a 1 in 'major_check', I'd like to find a 12121 pattern in 'minor_check'
I highlighted a 21212 pattern in the screenshot to give a better idea on what I am looking for.
Once the 21212 or 12121 patterns are found, I'll check if specific rules applied on open/high/low/close (corresponding to the 5 rows constituting the pattern) are met or not.
Of course, one could naively iterate through the dataframe but this doesn't sound like the Pythonic way to do it.
I didn't manage to find a good way to do this, since a 21212 pattern can have some 0s inside it
Источник: https://stackoverflow.com/questions/777 ... nding-rows
