Hi

Is there an indicator tutorial that gives a newbie a step by step guide about writing indicators ... I want to write an idicator for a tick chart that needs to look atthe last n ticks in order to calculatesome outputs ... so I am wondering about the performance aspects of this and how to do it!

And 1 specific question is it possible to output candle plot in a sub pane with some calculated values of H/L/O/C


Thanks for any help in advance ...


Paul

Browse through the examples:
https://www.dukascopy.com/wiki/#Simple_Indicator
For more examples see:
https://www.dukascopy.com/wiki/#Platform_indicator_source_files
Note that on tick charts the ticks get aggregated to 1 sec bars, hence you won't work with n-ticks rather 1-sec bars.
The indicator draws either on the main chart or on a subpanel - it can't do both. However, you can draw the OHLC values also on the main chart, see how the "Indicator calculates bar heights in pips" text gets plotted in the following example:
https://www.dukascopy.com/wiki/#Indicator_draws_chart_objects

If you wish to do "pure tick" processing, be aware that you can do those calculations in an IStrategy module. For example, you might want to do tick interpolation or "sparse tick" handling in calculations which an Indicator might make difficult.

Pure tick "spike analysis" is another thing you could do, using tick detail which is not aggregated into 1 second pseudo-bars, as would be the case with an indicator attempting such a calculation.

Also, you have access to Depth of Market data (not in backtesting or historical mode, only Live/Demo modes) which indicators may also not be able to get, I'm not a real expert on writing Indicators either

It is not necessary for your calculations to have a visual component, as Indicators provide, so just keep this in mind. And the IStrategy module can gain access to charts for purposes of placing visual indications, should that be desirable.

If you need any info on this, including Tick Queueing let me know. Tick Queueing is where you capture all the information associated with the onTick callback, including depth of market and timestamping, and then queue it up for later immediate processing.

This frees up the onTick callback thread to process further events while you process the timestamped queued data on other threads.

HyperScalper