OPEN-SOURCE SCRIPT

Multi-ZigZag Multi-Oscillator Trend Detector

已更新

This table is intended to give you snapshot of how price and oscillators are moving along with zigzag pivots.

This is done in the same lines of Zigzag-Trend-Divergence-Detector
But, here are the differences

Table shows multiple oscillator movements at a same time instead of one selected oscillator
Divergence is not calculated and also supertrend based trend. Trend can be calculated based on zigzag movements. However, lets keep this for future enhancements.
This system also uses multiple zigzags instead of just one.

⬜ Process

▶ Derive multiple zigzags - Code is taken from Multi-ZigZag
▶ Along with zigzags - also calculate different oscillators and attach it to zigzag pivot.
▶ Calculate directions of zigzag pivots and corresponding oscillators.
▶ Plot everything in the table on last bar.

⬜ Table components

Table contains following data:

Directional legends are:

⇈ - Higher High (Green)
⇊ - Lower Low (Red)
⭡- Lower High (Orange)
⭣ - Higher Low (Lime)

⬜ Input Parameters

▶ Source : Default is close. If Unchecked - uses high/low data for calculating pivots. Can also use external input such as OBV
▶ Stats : Gives option to select the depth of output (History) and also lets you chose text size and table position.
▶ Oscillators : Oscillator length is derived by multiplying multiplier to zigzag length. For example, for zigzag 5, with 4 as multiplier, all oscillators are calculated with length 20. But, same for zigzag 8 will be 32 and so on.
▶ Available oscillators :

CCI - Commodity Channel Index
CMO - Chande Momentum Oscillator
COG - Center Of Gravity
MFI - Money Flow Index (Shows only if volume is present)
MOM - Momentum oscillator
ROC - Rate Of Change
RSI - Relative Strength Index
TSI - Total Strength Index
WPR - William Percent R
BB - Bollinger Percent B
KC - Keltner Channel Percent K
DC - Donchian Channel Percent D
ADC - Adoptive Donchian Channel Percent D (Adoptive-Donchian-Channel)

⬜ Challenges

There are 12 oscillators and each zigzag has different length. Which means, there are 48 combinations of the ocillators.

First challenge was generating these values without creating lots of static initialization. Also, note, if the functions are not called on each bar, then they will not yield correct result. This is achieved through initializer function which runs on every bar and stores the oscillator values in an array which emulates multi dimensional array oscillator X zigzag length.

Next challenge was getting these values within function when we need it. While doing so I realized that values stored in array also have historical series and calling array.get will actully get you the entire series and not just the value. This is an important takeaway for me and this can be used for further complex implementations.

Thanks to LonesomeTheBlue and LucF for some timely suggestions and interesting technical discussions :)

發布通知

Major update

Added Zigzag based supertrend for all zigzag levels. Original implementation of zigzag supertrend is here: ZigZag-Supertrend
Added divergence sentiment to all the oscillators which is calculated based on the initial indicator. Detailed rule:
The script is pretty heavy and it started throwing pine error on number of lines exceeding 60000. To avoid this, had to remove some of the oscillators/indicators. Have removed Momentum which is quite different from most of them. And have removed Keltener channel, Donchian Channel and Advanced Donchian Channel as they produce almost similar results as that of Bollinger Band.

PS: If you need older version you can load the script to codebase and then select different version of it from version history.

發布通知

Major Updates

Redefined color coding to show different colors for divergence and hidden divergence.
Bullish and bearish indecision is marked with different symbols (⤒ and ⤓) wheras invalid divergence and hidden divergence preserves the existing indecision symbol - ▣

發布通知

Convert to pine 5