Constant Median Line across a lower study

[hbsopen]

I got this thought that; there are studies where you have a zero-line and then inputs for overbought ( = 80 usually) and oversold ( = 20 usually). But instead of such fixed values, why not have a Median (not a moving average) but a static value across for the Aggregation period and Time frame. I do understand that for a 20 year chart this may not make sense, but I intend to use it only for shorter periods like 5-days or 10-days..

Then, I will enhance such that, instead of overbought and oversold - I would have inputs for factor above and below that median price. This piece, I think, I can figure out looking at the code of some indicators where input is standard deviation (for e.g VWAP). What I am not being able to get to is the static Median line (instead of a zero line). I did try to use the Median* functions but those use price and for me it could be volume or it just could be an oscillator value or any other calculated value.

See below, so something similar to this but median value.

plot RS = if stddevVol == 0 then 0 else stddevVol;
RS.setDefaultColor(GetColor(6));

def sr = CompoundValue("historical data" = RS, "visible data" = if isNaN(sr[1]) then RS else sr[1]);

So, above generates a static value across the aggregation and time frame selected. But, I don't believe, from what I read in ThinkScript help, this result in a Median, But I think you get the point.

I did search 'Media' in thread titles and didn't find a result that would lead me to above.

 

[AndrewPnF]

I got this thought that; there are studies where you have a zero-line and then inputs for overbought ( = 80 usually) and oversold ( = 20 usually). But instead of such fixed values, why not have a Median (not a moving average) but a static value across for the Aggregation period and Time frame. I do understand that for a 20 year chart this may not make sense, but I intend to use it only for shorter periods like 5-days or 10-days..

Then, I will enhance such that, instead of overbought and oversold - I would have inputs for factor above and below that median price. This piece, I think, I can figure out looking at the code of some indicators where input is standard deviation (for e.g VWAP). What I am not being able to get to is the static Median line (instead of a zero line). I did try to use the Median* functions but those use price and for me it could be volume or it just could be an oscillator value or any other calculated value.

See below, so something similar to this but median value.

plot RS = if stddevVol == 0 then 0 else stddevVol;
RS.setDefaultColor(GetColor(6));

def sr = CompoundValue("historical data" = RS, "visible data" = if isNaN(sr[1]) then RS else sr[1]);

So, above generates a static value across the aggregation and time frame selected. But, I don't believe, from what I read in ThinkScript help, this result in a Median, But I think you get the point.

I did search 'Media' in thread titles and didn't find a result that would lead me to above.

Don't know what you really need but this seems to track the median value of an oscillator. Don't know if it provides additional "value" but it may be something you can play around with.

# Dynamic Median Line for Stochastic
declare lower;

input over_bought = 80;
input over_sold = 20;
input KPeriod = 14;
input DPeriod = 3;
input medianLength = 20;

plot FullK = StochasticFull(over_bought, over_sold, KPeriod, DPeriod, high, low, close, 3, AverageType.SIMPLE).FullK;
plot FullD = StochasticFull(over_bought, over_sold, KPeriod, DPeriod, high, low, close, 3, AverageType.SIMPLE).FullD;

# The Custom Median Line
plot StochasticMedian = Median(FullK, medianLength);
StochasticMedian.SetDefaultColor(Color.YELLOW);
StochasticMedian.SetStyle(Curve.SHORT_DASH);

plot Level80 = over_bought;
plot Level20 = over_sold;

 

[hbsopen]

Thanks.

I realized that 'constant' may not be mos accurate term. I said that because I don't day-trade (have full time work), so I come back and look at the market or on Friday/ weekend. So if I look when markets are closed, that line will be a constant. I do understand that the line will be dynamic during trading hours. A better term is 'flat line'.

Having said that it could be average flat line (but not like a moving average). So essentially there would be no 'input' for length, as in your suggestded code. Instead it would be an average (or median) for all the bars in the time-interval (5 m, 15 min .. or even daily) across the time frame (5 days or 10 days or whatever is selected). That line would be a flat line across the study.

Hoping this can be done.

 

[Joshua]

My guess is that you're looking for something like this, but you're not sure how to phrase it. I don't think the exact median value all that important in the context of what you've described. What you want sounds more like the center of the visible range.

def H = highestall(something);
def L = lowestall(something);
plot x = ((H - L) / 2) + L;