Directional Analyzer

2021년 7월 31일

This script attempts to equip users with the necessary information about the direction of an instrument, and essentially it is a synergy of 3 algorithms.

The first algorithm (plotted as dots at level 0) studies the balance of delta volatility that constitutes the current bar and answers if bulls or bears are in control at that exact bar time

The second algorithm (plotted as an area) studies the development of delta volatility over the defined period by means of a polynomial regression. Effectively, it provides an overall picture of the trend strength.

The third algorithm (plotted as a line with arrow labels) utilizes simple elements of neural network in conjunction with some custom filters to predict the focal point that a trend will reverse its direction. This is predictive in nature, hence always adopt this with caution. While the labels display the predicted direction, the colors of the line also reflect the state of the current bar as well, adding to the confirmation of the first algorithm.

May you be on the right side of the trade.

2021년 8월 26일

릴리즈 노트

Update the logics behind the predictive models.
Further clean up the code for faster execution.

2021년 8월 29일

릴리즈 노트

Major Update:
- A new predictive algorithm has been included, and the correlation between this new one and the old one will be utilized to provide directional forecast now plotted as arrows. The line part has been removed due to underperformance in giving confirmation of direction.

- A self-optimizing filter has been added to the algorithms that analyze the current bar state (plotted as dots at zero line) and the overall strength of the trend (plotted as the area - now made parameterless). This filter significantly removes the lag and overshoots compared to without the filter

- A new algorithm has also been added (plotted as columns), which is based on my interpretation of entropy in information theory. This shall perform quite similarly to the area part, but tailored to usually lead the overall strength of the trend. This mean that the confluence of the columns part and areas part gives extra conviction to trend bias.

- Dynamic Overbought/ Oversold levels are introduced to identify potential trend reversal (usage: compared against the area part)

2021년 8월 29일

릴리즈 노트

Minor changes in default settings

2021년 10월 5일

릴리즈 노트

Migration to Pine V5