What does it mean if one day's closing price is visible on another day? So what information does being visible give us? The graph can also be created by other methods. What is the logic of this method?
Very well explained. I look forward to more on similar topics. Have you considered structural entropy thresholds for risk-on risk off signals? There have been some other posts and videos on this but your excellent python coding and clear explanations would add to this. Tank you
Awe yeah. Great stuff. Would be super cool if you wrote a function of a Lagrangian and compared the mins and max’s Maybe see which one finds it faster ?
Does it really mean anything with trading data that a node is "visible" from another node. Intuitively I don't see the connection. Visibility graphs are normally used for objects in Euclidean space, right? cheers.
This is good for retrospective analysis only because you don't know how high or low the next bar will be. Unless there's a working theory on extrapolation solely based on visibility graphs. But still the future is not visible from the present so I doubt whether it is really possible. I have found derivative chain data to be highly reliable for first information
Interesting method, but I don't understand how we are supposed to use this technique. Could we identify cycles in the graph to find some sort of seasonality in the time series?
Very interesting video, thank you. I think I understand the idea behind using a visibility graph, but I'm not sure about the symmetry of that adjacency matrix. A point in the past is visible from the past, but not vice versa, so "visibility" in the "visibility graph" could be not quite true. It would be interesting to see one-directional links which I guess would make that graph a tree. However the shortest path is a metric available only in a fully connected graph, so I'd need to read about graph theory a lot to understand which metric would be a good candidate for this experiment.
I talk about the direction of 'visibility' in my video about time series reversibility. You can find the forward links by looking only at the upper triangle of the adjacency matrix.
![Felix "Unfair" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/Oswujxm2Ag0/mqdefault.jpg)
Every video you produce is more fascinating than the previous..
wow all the math i learned back 20 years ago suddenly make sense now, thanks bro !!
Thanks again for the amazing insights, knowledge, and techniques that you teach! Another banger video :)
Thank you! This feature will make an excellent addition to my collection.
Nice stepping through your code. Thanks for a very interesting topic I never heard of before
thank you, new ideas are always good, i personally enjoyed this video and looking forward to new ones...
What does it mean if one day's closing price is visible on another day? So what information does being visible give us? The graph can also be created by other methods. What is the logic of this method?
This is my kind of content!! Thank you!!!
Very well explained. I look forward to more on similar topics. Have you considered structural entropy thresholds for risk-on risk off signals? There have been some other posts and videos on this but your excellent python coding and clear explanations would add to this. Tank you
Fascinating video, keep up great work
Thanks a lot for the insightful video!
Awe yeah. Great stuff. Would be super cool if you wrote a function of a Lagrangian and compared the mins and max’s
Maybe see which one finds it faster ?
Does it really mean anything with trading data that a node is "visible" from another node. Intuitively I don't see the connection. Visibility graphs are normally used for objects in Euclidean space, right? cheers.
I like the look of the horizontal visibility graph. Maybe I can post a request for an indicator on MQL4? Wonder how much it would cost
This is good for retrospective analysis only because you don't know how high or low the next bar will be. Unless there's a working theory on extrapolation solely based on visibility graphs. But still the future is not visible from the present so I doubt whether it is really possible. I have found derivative chain data to be highly reliable for first information
as always greatfull for the knowledge that you share
something new to me, thanks!
Interesting method, but I don't understand how we are supposed to use this technique. Could we identify cycles in the graph to find some sort of seasonality in the time series?
Amazing content man, thanks a lot. What code editor are you using? vim?
Next level stuff amazing
it was very interesting, thank you ✌️
Very interesting video, thank you. I think I understand the idea behind using a visibility graph, but I'm not sure about the symmetry of that adjacency matrix. A point in the past is visible from the past, but not vice versa, so "visibility" in the "visibility graph" could be not quite true. It would be interesting to see one-directional links which I guess would make that graph a tree. However the shortest path is a metric available only in a fully connected graph, so I'd need to read about graph theory a lot to understand which metric would be a good candidate for this experiment.
I talk about the direction of 'visibility' in my video about time series reversibility. You can find the forward links by looking only at the upper triangle of the adjacency matrix.
Hey so much great content but if I were to follow I best which of you strategy would you recommend me to go with?
it looks like its measuring convexity
fancy but hard
I tried to reach out via email please check it out