Making a Python trading system for MT. - page 13
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
For Alex, introdynamics in python.
install version 3.7 64 bit (I don't use anaconda and don't understand why you need it, it's probably for too smart people)
open a command line and type pip install catboost
this will install the catboost and give you a warning about what libs are missing
Another option is to pip install jupyter notebook (pip install jupyter notebook) or jupyter lab
Googling for more details
Thank you, I will give it a try.
What's the name of the parameter enumeration package? Some kind of "grid"...
I don't know, I set it up from a YouTube video.
Another question, - can the current regression line be used to calculate the distribution? To do this, let us build 2 regression lines RegLine1 - from the end point of the plot to a depth of 800 min, and a similar RegLine1 from the point with a shift of 200 min backwards.
We obtain the graph:
We see that the old RegLine2 is almost identical to the new, freshly plotted RegLine1 on the latest data.
Note that this is not always the case, it can be worse, but the total errors are quite acceptable and much lower in comparison with other average simulation methods. However, this picture is also quite typical.
And, of course, the smaller the shift, the smaller the error. We kind of took the worst-case scenario to see the process of divergence over time.
There were a few tricks involved, but that will be left out of the picture).
Well done, Yuri!
You have taken up one of the fundamental questions that has not been well researched in T&C.
What is the measure of central tendency in a market process? What kind of average should the confidence interval be based on? Do I understand your research correctly?
Yes, yes, it's not the MA or the median. I personally use WMA with some weights. And your regression lines are quite interesting in that regard.
Or are you just demonstrating the capabilities of Python? Explain the purpose of your thread, please.
2 lines is not enough, 200 would be better.
I don't get it. What do you mean?
I don't get it. What are you talking about?
The little tricks.
about the little tricks of the trade.
That's another topic. Behind the scenes and no comments).
Above written -Note that this is not always the case, it can be worse, but the cumulative errors are acceptable ...Acceptable for my purposes.That's good enough. For some others, I don't know.
By the way, and the goals have been voiced -the cumulative errors are quite acceptable, and significantly less compared to other methods of simulating the mean.
In addition to my previous post, looked at the distribution relative to the regression line at 3000 counts. At shorter intervals it is very jagged.
Actually it is very unstable, and its shape changes a lot from plot to plot, but the min and max deviations remain at about the same levels. Well, and there's no trace of the long tails. I am not making any conclusions, see for yourself.
I can only say that distribution tails are a result of our actions and not a property of the market.
In addition to my previous post, looked at the distribution relative to the regression line at 3000 counts. At shorter intervals it is very jagged.
Actually it is very unstable, and its shape changes a lot from plot to plot, but the min and max deviations remain at about the same levels. Well, and there's no trace of the long tails. I am not making any conclusions, see for yourself.
I can only say that distribution tails are a result of our actions and not due to the market.
If the lines did not overdraw too much it would be fine.
what period and degree?