Scan Using Future Bars & Highest All

[chewie76]

I tried scanning for when the histogram is blue, like highlighted in the below picture. I created a plot for this called "blue". For some reason, the results are not working at all. What am I doing wrong?

Scan
http://tos.mx/!UInj9Whd

Indicator
http://tos.mx/!eQfxYyXo

 

[merryDay]

I tried scanning for when the histogram is blue, like highlighted in the below picture. I created a plot for this called "blue". For some reason, the results are not working at all. What am I doing wrong?

I don't use future bars or the deprecated HighestAll() function

def isLastBar = IsNaN(regAvg[-1]) and !IsNaN(regAvg);
def regCapture = if isLastBar then regAvg else Double.NaN;
def regFlat = HighestAll(regCapture);

I am just assuming that is where the problem lies.
Here are the issues:
Use of future bars means that the trigger is not painted on the current bar but repainted further back.
Therefore, adding within x bars with allow the scan to find results.

The bigger issue is the Highest All dependency makes the level inconsistent across scans/charts/watchlists because it always references the absolute last bar of the entire dataset, not a fixed lookback for scan, watchlists, or charts;
Given that the dataset available on your chart will never match the dataset in the scan; it will always result in inconsistencies. Additionally, Schwab deprecated the HighestAll() function resulting in it; not updating in real time.
Here is my workaround:
https://usethinkscript.com/threads/premarket-charts-analysis-followup.21879/page-3#post-158874

Here is the code resolving the forward looking bars BUT THE ISSUE with HighestALL errors remain.

#Williams_Vix_Fix_BollingerBand

input pd = 22;                     # Period for Highest Close
input length = 20;                 # SMA Length
def highestClose = Highest(close, pd);
def WVF = (highestClose - low) / highestClose * 100;

input buyline = 2.0;
input lookback = 400;

def regApprox = Inertia(WVF, length);
def regAvg = if BarNumber() > lookback
                then Average(regApprox, lookback)
                else regApprox;

def isLastBar = IsNaN(regAvg[-1]) and !IsNaN(regAvg);
def regCapture = if isLastBar then regAvg else Double.NaN;
def regFlat = HighestAll(regCapture);
def PercentLineUP = regFlat * buyline;



def blue = WVF > PercentLineUP and WVF < WVF[1] ;

plot scan = blue within 3 bars;

 

[Joshua]

It's working fine on my end... it might be the volume filter if you were working on it after hours. Daily volume resets back to zero at some point, not sure when exactly, but that filter will deny results if you have a minimum after the reset. Just a theory for now anyway, won't know for sure until later tonight.

 

[chewie76]

I don't use future bars or the deprecated HighestAll() function


I am just assuming that is where the problem lies.
Here are the issues:
Use of future bars means that the trigger is not painted on the current bar but repainted further back.
Therefore, adding within x bars with allow the scan to find results.

The bigger issue is the Highest All dependency makes the level inconsistent across scans/charts/watchlists because it always references the absolute last bar of the entire dataset, not a fixed lookback for scan, watchlists, or charts;
Given that the dataset available on your chart will never match the dataset in the scan; it will always result in inconsistencies. Additionally, Schwab deprecated the HighestAll() function resulting in it; not updating in real time.
Here is my workaround:
https://usethinkscript.com/threads/premarket-charts-analysis-followup.21879/page-3#post-158874

Here is the code resolving the forward looking bars BUT THE ISSUE with HighestALL errors remain.

#Williams_Vix_Fix_BollingerBand

input pd = 22;                     # Period for Highest Close
input length = 20;                 # SMA Length
def highestClose = Highest(close, pd);
def WVF = (highestClose - low) / highestClose * 100;

input buyline = 2.0;
input lookback = 400;

def regApprox = Inertia(WVF, length);
def regAvg = if BarNumber() > lookback
                then Average(regApprox, lookback)
                else regApprox;

def isLastBar = IsNaN(regAvg[-1]) and !IsNaN(regAvg);
def regCapture = if isLastBar then regAvg else Double.NaN;
def regFlat = HighestAll(regCapture);
def PercentLineUP = regFlat * buyline;



def blue = WVF > PercentLineUP and WVF < WVF[1] ;

plot scan = blue within 3 bars;

I appreciate your efforts. From what I'm showing as results, it doesn't look like it's capable. Thanks for checking.