Logging the Time of an Alert for Future Reference

[TNTrader5159]

To determine which end of an engulfing bar surpassed the minimum number of bars required first, is it possible to hold the time one end of the bar surpasses them in queue (as with the time of an alert upon that condition being met), and upon the other end of the bar surpassing the minimum number of bars in the other direction to compare the two times and determine which came first intra-bar?

For example, with a 3-bar minimum (reflecting an overall minimum ratio of 3:1 for bars vs prior bars engulfed), the high of a 5-minute bar surpasses the 3 previous bars' highs at 11:30:36 AM. Then at 11:32:03 the low of that same 5-minute bar surpasses the 3 previous bars' lows. Upon comparing the two times, you know which came first. Saying the open must be higher than the close works most of the time but not always.

A possible answer would be to use the two most recent bars and have the previous one surpass the highs of the minimum 6 previous bars (same minimum 3:1 ratio) and the current bar surpass the lows of the 7 bars before it that would include the previous bar.

That solves the problem and still allows the chart to "turn over" every 5 minutes for better accuracy than with using 10-minute bars. You effectively trade on a 10-minute time-frame using 5-minute bars here.

Also we must consider that if the previous bar engulfs said range on its own and the current bar is inside it, we have a better ratio that would loosely be double that of the 2 bars engulfing the previous 6 or more bars basis this indicator without provision for that possibility.

This seems the way to go for best accuracy, if the time-alert method outlined above is not feasible.

What say y'all?

 

[halcyonguy]

To determine which end of an engulfing bar surpassed the minimum number of bars required first, is it possible to hold the time one end of the bar surpasses them in queue (as with the time of an alert upon that condition being met), and upon the other end of the bar surpassing the minimum number of bars in the other direction to compare the two times and determine which came first intra-bar?

For example, with a 3-bar minimum (reflecting an overall minimum ratio of 3:1 for bars vs prior bars engulfed), the high of a 5-minute bar surpasses the 3 previous bars' highs at 11:30:36 AM. Then at 11:32:03 the low of that same 5-minute bar surpasses the 3 previous bars' lows. Upon comparing the two times, you know which came first. Saying the open must be higher than the close works most of the time but not always.

A possible answer would be to use the two most recent bars and have the previous one surpass the highs of the minimum 6 previous bars (same minimum 3:1 ratio) and the current bar surpass the lows of the 7 bars before it that would include the previous bar.

That solves the problem and still allows the chart to "turn over" every 5 minutes for better accuracy than with using 10-minute bars. You effectively trade on a 10-minute time-frame using 5-minute bars here.

Also we must consider that if the previous bar engulfs said range on its own and the current bar is inside it, we have a better ratio that would loosely be double that of the 2 bars engulfing the previous 6 or more bars basis this indicator without provision for that possibility.

This seems the way to go for best accuracy, if the time-alert method outlined above is not feasible.

What say y'all?

no, can not record any signals, during a bar, except low and high. all variable values will fluctuate during the current bar, but only the values that exist when the bar closes, will be able to be read, by future bars.

 

[TNTrader5159]

no, can not record any signals, during a bar, except low and high. all variable values will fluctuate during the current bar, but only the values that exist when the bar closes, will be able to be read, by future bars.

Thanks for the clarification, so I'll work around the inexactitude. I wonder if TD Ameritrade would consider developing a comparative time stamp indicator for the highs and lows or other facets of a live or mature bar. I guess I could ask huh.

 

[halcyonguy]

To determine which end of an engulfing bar surpassed the minimum number of bars required first, is it possible to hold the time one end of the bar surpasses them in queue (as with the time of an alert upon that condition being met), and upon the other end of the bar surpassing the minimum number of bars in the other direction to compare the two times and determine which came first intra-bar?

if your desired timeframe is greater than 1 minute, like day or hour, then a smaller time (like 1 minute) could be used to gather data.

set chart at 5 minute.
in a study, use 2nd aggregation, set to the longer time, to gather your primary data.
then use chart data to 'evaluate' the bars that occur during the longer time.

example,
chart - 5 minute
2nd agg set to 60 minute , results in 60/5 = 12 chart bars per longer (agg) time. evaluate the highs and lows , (or some desired variable) during each group of bars.

 

[TNTrader5159]

if your desired timeframe is greater than 1 minute, like day or hour, then a smaller time (like 1 minute) could be used to gather data.

set chart at 5 minute.
in a study, use 2nd aggregation, set to the longer time, to gather your primary data.
then use chart data to 'evaluate' the bars that occur during the longer time.

example,
chart - 5 minute
2nd agg set to 60 minute , results in 60/5 = 12 chart bars per longer (agg) time. evaluate the highs and lows , (or some desired variable) during each group of bars.

Yes, I was planning to do just that for time-frames of 2 minutes or greater so it will be most accurate, eg 2 bars must engulf at least 6, 3 engulf 9, 4 engulf 12, etc. For a single bar, I can do as you have done just using close > open or vice versa that will make it right most of the time on historical charts and accurate for real-time alerts as long as close is used for the trigger point as the price crosses the engulfed range to complete the pattern. For say two bars and up, the last bar must be the trigger bar, so that will clearly identify which end came first high or low. If using say 10 bars engulfing 30, if the "engulfment" takes place in an earlier bar within those 10, it would show up on the 9/27 ratio or 8/24 etc. So examining all timeframes from 1-minute on up as high as you wanna go is the way to find the most and most accurate trade opportunities, as long as it doesn't consume too many computing resources.

So all that said, thanks a billion for your guidance! I have begun coding my system a segment at a time and get excited when no red enshrouds my code and it produces the results I want.