From theory to practice - page 1439

 
me
Alexander_K:

Gentlemen!!!

Zhenya is silent... And I, edge of the shout, need test results on various currency pairs.

Does anyone here know how to do quick EAs, and knows what a RMS is?

I'm ready to repeat Koldun's algorithm again.

It is not my algorithm, the author has given his permission for publication, and if it is profitable, everyone may use it.

My personal request - I need the test results by tomorrow.

I don't give a shit, I have plenty of time, I even remake other people's robots for fun) What do you want, Sash? I have a ready-made SKO).

but let's not... um... "weird" thoughts please, no bullshit)
 
Martin_Apis_Bot Cheguevara:
me

I do not care I have plenty of time for nothing, I even do other people's robots) just for fun) What do you need Sash? I have a ready-made SKO)

without... ... "weird" thoughts please, no bullshit)

For GBPUSD, December 2018 - March 2019 inclusive, you should get the following chart:

There should be 12 trades. 10 to the plus side, 2 to the minus side. A total profit of 300 pips.

If it's not working, then you're doing something wrong. Either it's all about my data - I don't have OPEN M1, after all, but my own thinned data.

 
Martin_Apis_Bot Cheguevara:

I don't have time to do more than that on GBPUSD...

Need tests on other pairs and over a longer period of time. Thank you!

 
Alexander_K:

For GBPUSD, for December 2018 to March 2019 inclusive, you should get this chart:

There should be 12 trades. 10 to the plus side, 2 to the minus side. A total profit of 300 pips.

If it's not working, then you're doing something wrong. Or it may be because of my data - my data is not OPEN M1, I have my own thinned data.

Anyway, I've read about the above-mentioned RMS

The Warlock has a more correct graph from a statistical point of view

his red and blue are smoother, so he takes the window to observe more than you do, I suspect he puts most of the history into the statistical formulas, this is so that the distribution is more or less close to normal

you don't have a lot of data to estimate, as I understand it
 
Renat Akhtyamov:

Anyway, I've been reading up on the proverbial RMS.

The sorcerer's graph is more statistically correct.

his red and blue are smoother, so he takes a larger window to observe than you, I suspect that he puts almost all the history into the statistical formulas, this is so that the distribution is more or less close to normal

you don't have a lot of data to estimate, as I understand it

You're asleep, aren't you? :)))

 
Alexander_K:

You're asleep, aren't you? :)))

You'll fall asleep in here.

I can't get past the things that at least two people are talking about together, and they're the ones with brains in their heads.

 
Renat Akhtyamov:

Anyway, I've been reading up on the proverbial RMS.

Wizard has a better graph from a statistical point of view

his red and blue are smoother, so he takes a larger window to observe than you do, I suspect he puts almost the whole history into his statistical formulas, it's to make the distribution more or less close to the normal

You have not enough data to make an assessment, as I understand it

He either has a larger window, or in the window =day or week works with ticks.

 
Alexander_K:

It either has a larger window, or the window =day or week works with ticks.

a naked eye look at the statistics:

ticks per day = large number

in calculating the deviation divided by the number = small

conclusion: we do not see big ones, plus they are discarded beyond 3 sigmas

Well I kind of just read it, maybe it's just a guess...

 

Who is good at coding on mt4 ?

Need to get an interesting EA to work on the latest build of mt4.

Please send a message to

 
Alexander_K:

:))) Are you a trader or a chekist? I don't get it...

I asked for a quick one, until tomorrow. It could take me a year to set up my own tester in WisCime.

Okay, forget it...

VisSim and other RAD toys are quite a slippery slope, the same goes for modern trends with python/PR and their 100500 libs, everything is licked to shine on presentation, useless examples, where something is complicated in 3 lines, and what isn't written in C++ and wrapped in a ready-made class with nice methods, it's impossible to fetch, at best, if realizable at all. RAD is good when you can EASILY integrate your "cubes" in a normal C-like language, if it's absent or very cumbersome, it will not do any good. So get your act together and learn C++, then write a tester in a couple of hours that will run strategies for years of ticks in seconds.