thread: Recent Strength indicator

Can it be possible to have a Relative Strength indior that has a specific start time as the bottom line and the NumPoints becomes the equal number of branches since that start time base line.

I'll take a look at this at any future point. Not sure when, nevertheless it is not a task and I'm really snowed under with other work at the moment.

Hello David, mooselover, isn't this only a naming/labeling matter? ....

I haven't tried what you explained. However, if it works, then good!

I'll take a look at this as at any future stage. Not sure I am really snowed under with other programming work right now and when, nevertheless, as it's not a trivial job.

Ok, Thank you for this...

I'll take a look at this at some future point. Not certain when, however it's not a trivial job and I am really snowed under with additional work right now.

I haven't tried what you described. But if it works, then great!

Hi Hanover,

Further to my previous Qs and correct me if I am wrong....

But would a fixed punctually BasePoint permit you to have, that if index graphs of each currency crossed they could quite easily be a good trade for the specified pair???

Cause a fixed base point will eliminate 2 continuing changing factors... The factors of beginning time, and beginning value of currencies such as index calculation, would become stable user input factors.

Due

Hi Hanover,

Further to my previous Qs and correct me if I'm wrong....

Not quite sure what you mean by fixed, and the potential impliions. The way the indy currently works, if you set the BasePoint to a'fixed' value, e.g. to 6 to a H1 chart, then the traces will probably disperse from an'origin' value that is always 6 hours ago, i.e. 6 hours old than the currently forming candles on the hourly price charts that are being aggregated. To correct the BasePoint at a point in time, e.g. MT4 midnight, will need me to alter the code. If (for the sake of backward compatibility with the existing concept) I had been to retain the current paradigm, that would effectively mean having another BasePoint counter for every phase, e.g. if the time is currently MT4 4 a.m., then that will be 4 candles after midnight a H1 plot; 8 candles after midnight on a M30 plot; 48 candles on a M5 plot; etc. What happens if the user chooses only 30 candles for the plot? That is one major reason designing the programming change would take some thought (let alone coding it).

But could a fixed punctually BasePoint allow you to have, that whenever index charts of each currency crossed they could very easily be a great exchange for the given pair???

Obviously I can't say if it's the crossover could be a good transaction any more than the MA crossover on a price chart might bring about a profitable commerce. But I assume that is not what you intended?

In any circumstance, a Recent Strength crossover in itself is not necessarily important. It merely implies that one currency's strength is rising and the other falling, and the crossover occurs only because of the way each one the values being plotted are being rescaled to fit a 0....100 range. For instance, if you were to eliminate some of the currencies in the plot, a rescale of those rest of the lines could occur, as well as the two currencies previously crossing over might no longer do. Likewise with changing any of the parameters.

Assuming that everything else is equal, the'greatest' commerce is obviously the pair that is the product of steepest rising line against the steepest falling line. But of course the selection of the origin point (where the steepness is being measured) is necessarily random. And it presumes that current strength is the basis for entry, rather than overboughtness/oversoldness (if a person is seeking to exchange'pullbacks', by way of example).

Cause a fixed base point will eliminate 2 continuing changing factors... The variables of starting time, and starting value of currencies for index calculation, would turn into stable consumer input variables.

To be pedantic, the values will always change because of the requirement to always rescale the plot to fit between a 0....100 range. But you're correct, this would mean any crossovers would stay fixed in the past, for any given set of parameters.... At least as far as I can imagine this, and assuming I understand exactly what you mean.

Given the way that the indy currently works, what you're describing is effectively the same as a'repainting' problem with other indys. However, like other repainters (e.g. a linear regression station ), although it impacts visual backtesting, it doesn't invalidate the meaningfulness of their value(s) being plotted in real time at the ideal hand border of the chart.

Mended was meaning a beginning point for calculations.

That's one major reason why designing the programming change would require some thought (let alone coding it).

Yes it is a very major reason and that I my remedial programming abilities had not thought of that it is better if timeframes used match up to available MT4 chart timeframes.

Obviously I can not say whether a crossover could be a fantastic transaction any more than the MA crossover on a price chart might result in a profitable commerce. However, I assume that is not what you intended?

Easily in my original statement was similar to significance possibly so yeah I was really meaning that...

Assuming that everything else is equal, the'best' commerce is obviously the pair that is the item of steepest rising line contrary to the steepest falling lineup. However, of course the choice of the origin point (where the steepness has been measured) is necessarily arbitrary.

This is my reason for wanting to have an unchanging beginning time, so the arbitrary steepness is removed.

the values will always change due to the need to continuously rescale the plot to fit between a 0....100 range. But you are correct, this would indicate any crossovers would remain fixed previously, for any given set of parameters.... At least as far as I can imagine it, and assuming I understand what you mean.

Thanks for answering my concerns on miniature details....


Its just that I'm looking at basket trading of currencies, and this indior of yours may be quite helpful for most basket trading procedures.

Thanks

greetings /david,

if it isn't too much to ask, it could be helpful if you can actually form the currencies into their most recent rank for all the chart. Besides the colour coding you've created, it would be a lot more easy for the eye to distinguish the traces.

I've attached a screenie with regard. I've drawn the lines to match the lines and also the currency abbreviations into illue my stage to our current settings.

I trust you find my suggestion useful and wouldn't complie the code too much.

Thank you and good day.

when it is not too much to ask, it'd be helpful if you can actually sort the currencies to their most recent rank for all the chart.

I will take a look at this as at any future point. Not certain when, however, as I'm really snowed under with programming work right now.

I assume that you understand that you can alter the colors and styles (they're parameter settings) to values which may be more significant to you?

Yes, I'm conscious of the colors that may be customized. It's just that I find myself, nevertheless matching the currencies to that colour I set, and my eyes moving out of the lines plotted to the currency to the right, like the (light yellow) lines I have drawn in my previous illuion.

Thanks again, and God bless you.

V 1.09 attempts to address most of the recent requests:

1) In the parameter file, colors may now rather be entered from the form (e.g.) r255b0g127 i.e. defining a red, blue and green value (0-255) for each primary colour in the desired mix.

2) New parameters LegendValues and LegendVertSpacing now optionally allow a color-coded screen of plot values at any specified point, and their relevant currency, on every period. See attached screenshot.

3) Instead of a rolling window of NumPoints candles, a fixed starting point may optionally be specified, for every period. This means that the window will grow as more data points are added. The starting point may be specified as YYYY.MM.DD HH:MM, or as a timeframe identifier (M5, M15, etc), which means that the window will automatically reset itself at the onset of each such timeframe.

4) Improved max NumPoints SmoothingSMA worth from 500 to 2,000.

I have not managed to test stage (3) fully because the market is currently shut. Please feel free to post if you have any issues.

See article #1 in this thread for your download files, also for detailed operational instructions.

kk