TODO

add test module to test against talib and/or pandas_ta.
can we have better implementations of SMA and SD, currently computing on whole data.
convert candle inputs to open, high, low and close.
create documentation.