is anybody using Zorro`s Frechet
code successfully for pattern recognition? I assume most of you are familiar with the article on Ronothwealth
The reason why I am asking is simple: the Zorro`s Frechet code uses only one Data Series (either close, high, low etc). But the problem is basically all patterns are defined by 2 lines: one connecting the highs and the other connecting the lows. Using close for Data Series is not a good solution for me since you cannot get good pattern detection because you are ignoring wicks.
Theoretically one could manipulate the Frechet`s code TimeFrame input (number of bars in the series to be compared). You could then try to detect a triangle visible on 4h bars by using huge number of 15 min bars & their close.
Anyway somehow I feel it would be great to have modified Frechet`s code, where we could define 2 separete series (one for highs and the other for lows). That would improve any pattern detection in my humble view.
Any thoughts are mostly welcomed. I am afraid the Zorro`s Frechet code cannot be "easily" modified since it is not open.