Entropy

[Crom]

https://www.hindawi.com/journals/complexity/2019/1817248/

https://www.hindawi.com/journals/complexity/2019/1817248/

I've been reading into some things and had an idea far less complicated than this. I wonder if there's a way to create a mathematical model than defines the entropy of the stock market at a given time period. Essentially you would use the level of entropy to determine the level of chaos/probability in the market.

 

[mashume]

https://www.hindawi.com/journals/complexity/2019/1817248/

https://www.hindawi.com/journals/complexity/2019/1817248/

I've been reading into some things and had an idea far less complicated than this. I wonder if there's a way to create a mathematical model than defines the entropy of the stock market at a given time period. Essentially you would use the level of entropy to determine the level of chaos/probability in the market.

While it is undoubtedly possible, there is a very real possibility that the mathematics involved is beyond the ability of ThinkOrSwim to model in real, or near-real time in any effective way -- if it is possible to compute in ToS at all that is.

That said, there is a possibility that an indicator could be coaxed out of data outside of ToS that could be used inside ToS. I've done this with the next-hour statistical indicator I created. It is a python script that does all the statistical calculations and spits out coefficients to a ToS indicator script.

This makes for interesting reading. Thank you. I'll give it a bit of thought. I enjoy these kinds of things.

-mashume

 

[rengagopal]

While it is undoubtedly possible, there is a very real possibility that the mathematics involved is beyond the ability of ThinkOrSwim to model in real, or near-real time in any effective way -- if it is possible to compute in ToS at all that is.

That said, there is a possibility that an indicator could be coaxed out of data outside of ToS that could be used inside ToS. I've done this with the next-hour statistical indicator I created. It is a python script that does all the statistical calculations and spits out coefficients to a ToS indicator script.

This makes for interesting reading. Thank you. I'll give it a bit of thought. I enjoy these kinds of things.

-mashume

mashume:- Could you please share how to bring data from outside python or c++ scripts into tos? Just a guide with example. Thank you!!

 

[mashume]

Here are two links to posts in a thread.

https://usethinkscript.com/threads/...ut-prediction-for-thinkorswim.6282/post-61282

https://usethinkscript.com/threads/...ut-prediction-for-thinkorswim.6282/post-61124

The first has a link to a notebook hosted in google collab which generates a ThinkScript indicator script. That code is hand-copied into the indicator code window in ToS.

The data is not used directly in ToS as data. It is massaged with python to become the code for the ToS indicator.

If I can explain more, do ask. I'll see how I can help.

-mashume

 

[rengagopal]

Here are two links to posts in a thread.

https://usethinkscript.com/threads/...ut-prediction-for-thinkorswim.6282/post-61282

https://usethinkscript.com/threads/...ut-prediction-for-thinkorswim.6282/post-61124

The first has a link to a notebook hosted in google collab which generates a ThinkScript indicator script. That code is hand-copied into the indicator code window in ToS.

The data is not used directly in ToS as data. It is massaged with python to become the code for the ToS indicator.

If I can explain more, do ask. I'll see how I can help.

-mashume

Thank you!!

 

[Crom]

While it is undoubtedly possible, there is a very real possibility that the mathematics involved is beyond the ability of ThinkOrSwim to model in real, or near-real time in any effective way -- if it is possible to compute in ToS at all that is.

That said, there is a possibility that an indicator could be coaxed out of data outside of ToS that could be used inside ToS. I've done this with the next-hour statistical indicator I created. It is a python script that does all the statistical calculations and spits out coefficients to a ToS indicator script.

This makes for interesting reading. Thank you. I'll give it a bit of thought. I enjoy these kinds of things.

-mashume

Thanks! That’s very interesting. I’d like to learn the basics of all this math and science so that I could start creating indicators based on this.

Do you have any suggestions of where I can start?

 

[mashume]

Thanks! That’s very interesting. I’d like to learn the basics of all this math and science so that I could start creating indicators based on this.

Do you have any suggestions of where I can start?

wikipedia is your best friend (and worst teacher / rabbit hole of knowledge and despair)

In all seriousness, I took a bunch of years of physics and mathematics and ended up teaching physics for a few years. I've also been programming computers and bending them to my will since some time in the early 1980s.

The maths and science are not really that difficult, once you have some basic handle on how mathematical proofs are built... but they can (are designed by authors?!?) to be intimidating.

On of the ways to learn how to program a complex function is to look for a function you are interested in and see if you can find a FORTRAN implementation of it. Fortran was a great scientific language back in the day (still used extensively in some sciences as it's well proven) and a very readable source code base. People tended to write concise fortran code that didn't make a bunch of imported function / package calls or include an object-oriented garbage and so it is simpler to see what the code is doing.

not sure how much concrete stuff there is in that brain dump to help you on your way, but feel free to write back with things you're working on.

-mashume

 

[Crom]

This is great, thanks! I shall do so.

Currently I am working on creating an indicator based on hurst cycles based on a subjective interpretation of the market. After that I hope to combine that with teachings based on Mandelbrot. I am very far from achieving these goals though, I only know simple math but in these past few months I have become extremely interested in science and thinkscript.