I don't get the benefits of using TA-Lib abstract ¿is it speed? ¿less processing? I went through the documentation, examples, and code examples, but can't get it. ¿can anyone explain it?
EDIT: This question is not opinion-based, as you can see in the answer, it's about performance of the function API vs abstract API of ta-lib
This API allows you to list all TA functions implemented in DLL (because python's ta-lib module is just a wrapper around compiled C library) and call them by indicator name (a string), instead of hardcoding switch with 200+ cases and rebuilding your app every time TA-lib adds new indicator. It also allows you to get info about data that indicator requires which may be used to adjust UI for end user asking him to enter the required data. So this API is designed not for a python data scientists who play with single TA indicator, but for desktop and web programmers writing GUI applications or web services addressed to wide audience. There is no any performance benefits.