Machine learning in trading: theory, models, practice and algo-trading - page 2847

 
Aleksey Nikolayev #:

It's not a histogram, or a histogram in the conventional sense as Pearson invented it.

Would it have made more sense to you if I'd written"bar charts"?

The generally accepted meaning in narrow circles, I should have specified....

Let's get to the heart of the matter :)

 
Aleksey Vyazmikin #:

Would it make more sense to you if I wrote"bar charts"?

No of course not, because it doesn't look like a bar chart either - the spacing between bars is uneven. If some bars are close together, they usually have a different colour and mean a different value to which the comparison is made.

Aleksey Vyazmikin #:

Commonly accepted meaning in narrow circles, it should have been specified.....

The circle of you alone is wide only from your point of view.

Aleksey Vyazmikin #:

Let's get to the substance of the question :)

The essence of your question so far lies only in its incomprehensible and incorrect formulation.

 
Aleksey Nikolayev #:

Of course not, because it doesn't look like a bar chart either - the spacing between bars is uneven. If some bars are close together, they usually have a different colour and mean a different value to which the comparison is made.

Well, what do you propose to call such a graph? Maybe a spectrogram then?

Aleksey Nikolayev #:

The circle of you alone is wide only from your point of view.

Do you want to make a study on this topic? Open excel and read what a histogram is for a wide range of users.

Aleksey Nikolayev #:

The essence of your question so far lies only in its incomprehensible and incorrect formulation.

What is not clear - what are predictors or ways of describing graphs?

 
Aleksey Vyazmikin #:

Well, what do you suggest we call such a graph? How about a spectrogram?

Do you want to do some research on this topic? Open excel and read what a histogram is for the general public.

What is not clear - what are predictors or ways to describe graphs?

At first glance, you have two values displayed that are not particularly related. In such cases, it is common to use a scatter plot.

 
Aleksey Nikolayev #:

At first glance, you have two values displayed that are not particularly related. In such cases, a scatter diagram is usually used.

We assume that the value is the same - just look at the red one. Green - the predictor response number is "1" and is always 1 for y.

 
Aleksey Vyazmikin #:

We consider that the value is the same - just look at the red one. Green - the predictor response number is "1" and is always 1 for y.

Each bar has an X coordinate of base and a Y coordinate of height, so there are two magnitudes. Since they are quite chaotic, you can look at their distribution on a plane - a heat map or something.

 
Aleksey Nikolayev #:

Each bar has an X coordinate of base and a Y coordinate of height, so two values. Since they are quite chaotic, you can look at their distribution on a plane - a heat map or something like that.

The x is the ordinal number of the observation, so scattering won't do anything but rotate the graph. What is important here, in my opinion, is the observation in dynamics.

 
In the first graph you can see the outliers and rows are quite similar to each other - some fence and trees. How to calculate the percentage of tree-emissions? Draw a line as close as possible to the edge of the fence?
 
Aleksey Vyazmikin first graph you can see the outliers and rows are quite similar to each other - some fence and trees. How to calculate the percentage of tree-emissions? Draw a line as close as possible to the edge of the fence?

Try to construct a histogram (in the usual conventional sense) for a sample of column heights. You could also try constructing a survival function.

 
Aleksey Nikolayev #:

Try to construct a histogram (in the usual conventional sense) for a sample of column heights. You can also try to construct a survival function.

I have built it, and how do you propose to use it?

As for the survival function, I don't understand how to use it.

Reason: