centre ville

Does anyone knows how to carry out standard deviation in jitter gen or pix codebox?

I realise I have to take the mean, however, how it's unclear to me how to accomplish this in jitter?

Hi,
Does anyone knows how to carry out standard deviation in jitter gen or pix codebox?
I realise I have to take the mean, however, how it's

standard-deviation-jitter-gen-codebox

 makes pretty good sense. No codeboxing necessary - there's nothing procedural about the calculations.

Since you'll need to keep running totals from sample to sample, adding some history to your patch will be your friend with the 

Doing it in jit.gl.pix makes pretty good sense. No codeboxing necessary - there's nothing procedural about the calculations.

Since you'll need to keep running totals from sample to sample, adding some history to your patch will be your friend with the jit.gl.pix object.

I'm guessing the sample or nearest operator could be used for this., or to feedback a slab to get some history - however, there is a lack of documentation to show how to do this for my use case.

I'm wondering what would also be the "data points" in jitter? Is this the number of pixels?

Jitter seems to use different/cryptic methods than other gfx programs and due to a lack of documentation on these, it may take weeks (and therefore be unfeasible) to achieve what would simply take a few minutes in other gfx programs.

 I'm guessing the sample or nearest operator could be used for this., or to feedback a slab to get some history - however, there is a lack of documentation to show how to do this for my use case. 

Here is what I mean about the documentation.

http://cycling74.s3.amazonaws.com/download/gen.documentation.beta.pdf

Sampling Sampling operators are one of the most powerful features of Jitter Gen patchers. Sampling operators take an input and a coordinate in the range [0, 1] as an argument, returning the data at the coordinate’s position in the matrix or texture. The first argument always has to be a Gen patcher input while the second argument is an N-dimensional vector whose size is equal to the dimensionality of the input it is processing.

I'm wondering now what's an N-dimensional vector? Surely if this is the same size as the iniput, then it is not N dimension but the same dimension as the input?

The next though moves to "If I am to calculate the mean, what do I calculate it on?" A vector? The whole matrix? A sample?

I get that it's powerful, however this doesn't explain how the operator is used. I'm wondering if anyone could decrypt this? Are there any gen pix codebox examples of using the "sample" or "nearest" operators, to show exactly what these do? These are operators that are not found in any other software and aren't really documented enough to be able to make use of them, for someone who isnt a guru.

Also, to find codebox examples is really hard. I'm actually searching through text on disk to find them. Would be great if there is some definitive list of ways to do things in codebox, or broken out examples of codebox.

 Sampling Sampling operators are one of the most powerful features of Jitter Gen patchers. Sampling operators take an input and a coordinate in the range [0, 1] as an argument, returning the data at the coordinate’s position in the matrix or texture. The first argument always has to be a Gen patcher input while the second argument is an N-dimensional vector whose size is equal to the dimensionality of the input it is processing. 

The next though moves to "If I am to calculate the mean, what do I calculate it on?" A vector? The whole matrix? A sample?

I get that it's powerful, however this doesn't explain how the operator is used. I'm wondering if anyone could decrypt this? Are there any gen pix codebox examples of using the "sample" or "nearest" operators, to show exactly what these do? These are operators that are not found in any other software and aren't really documented enough to be able to make use of them, for someone who isnt a guru.

https://cycling74.com/tutorials/winter's-day-jit-gl-pix-a-procedural-drawing-homage

 operator unless you're doing something procedural or engaging in some specific kinds of calculation (and I don't think you are doing that here). It doesn't buy you anything in terms of efficiency in this case, near as I can tell.

https://cycling74.com/tutorials/gen~-for-beginners-part-5-the-codebox-operator

This tutorial covers how the sample operator does its job reasonably well.

You really don't need a codebox operator unless you're doing something procedural or engaging in some specific kinds of calculation (and I don't think you are doing that here). It doesn't buy you anything in terms of efficiency in this case, near as I can tell.

https://cycling74.com/tutorials/gen~-for-beginners-part-5-the-codebox-operator 

if you're goal is to calculate the standard deviation of every pixel in a matrix or texture, then gen (jit.gen or jit.gl.pix) is not the right tool for this job. as described in your other thread gen's purpose is to apply your gen patch program to every pixel of its input, and write something to every pixel of its output.

you can design calculations that sample across the inputs (using sample or nearest operators) but you wouldn't design a program that samples every pixel for every output pixel.

if you're goal is to calculate the standard deviation of every pixel in a matrix or texture, then gen (jit.gen or jit.gl.pix) is not the right tool for this job. as described in your other thread gen's purpose is to apply your gen patch program to every pixel of its input, and write something to every pixel of its output. 

Hello CENTRE VILLE, can you post a draft patcher of what you want to do? 

It is unclear to me from which data (or dimension) you want to calculate the standard deviation and what is your processing speed requirement.

I think if you are accustomed to programming in Processing or Matlab where you have free access to all values of a matrix or array type the parallel processor objects like jit.gen or jit.gl.pix (fragment shader) might appear confusing. The sample object in Max has equivalents in shader languages like GLSL. In jit.gl.pix (GPU) it can be used to fetch arbitrary pixel values for calculation of the current fragment- and keep in mind all the code is executed in parallel for all fragments. This is e.g. useful for implementation of filters like Gauss or Sobel filters (see e.g. cf.sobel.jxs shipped with Max) - but it will not be a suitable approach for calculation of e.g. the mean value or standard deviation of the entire image. 

Hello CENTRE VILLE, can you post a draft patcher of what you want to do? 
It is unclear to me from which data (or dimension) you want to calculate the standard deviation and what is your processing speed requirement.
I think if you are accustomed to programming in Processing or Matlab where you have free access to all values of a matrix or array type the parallel processor objects like jit.gen or jit.gl.pix (fragment shader) might appear confusing. The sample object in Max has equivalents in shader languages like GLSL. In jit.gl.pix (GPU) it can be used to fetch arbitrary pixel values for calculation of the current fragment- and keep in mind all the code is executed in parallel for all fragments. This is e.g. useful for implementation of filters like Gauss or Sobel filters (see e.g. cf.sobel.jxs shipped with Max)  - but it will not be a suitable approach for calculation of e.g. the mean value or standard deviation of the entire image. 
 

I'm new in the gen world, so I'm still figuring out basic stuff.

I looked around in the forum and in the documentation, but I didn't find the answers I'm looking for.

image.png

This info doesn't mean much for a beginner without some basic example of usage.

I have a basic example where I have a one plane input matrix of size 5x2.

I expected to get that info from the dim operator but I only get 5's (the x axis).

I already tried all possible combinations the I could remember.

Eventually I wish to learn how to access this info in the codebox.

I know it's a bit off topic, but it's a tiny step of something bigger that I'm trying to achieve.

Hello Gen Jitters,
I'm new in the gen world, so I'm still figuring out basic stuff.
I looked around in the forum and in the documentation, but I didn't find the answers I'm looking for.

This info doesn't mean much for a beginner without some  basic example of usage.
I have a basic example where I have a one plane input matrix of size 5x2.
I expected to get that info from the dim operator but I only get 5's (the x axis).

I already tried all possible combinations the I could remember.
Does anyone know how explain this to me?

Eventually I wish to learn how to access this info in the codebox.
I know it's a bit off topic, but it's a tiny step of something bigger that I'm trying to achieve.
Thank you all!

Standard deviation jitter gen codebox?