Krunal

I have a PS4 controller connected to Max. There is a touch interface which allows me to determine whether I am swiping left or right.

Swiping results in an integer output ranging from 0-255, and swiping left/right leads to a derement/increment (respectively). A particularly quick/long swipe may result in the integer wrapping around before continuing to decrement/increment).

I have attached the method I have implemented so far, which does the job except at the boundary points.

Once the boundary of the integer range is reached, the numbers wrap around, which results in the wrong bang being triggered, for a single instance. e.g. a transition from 255 to 0 results in the left bang being triggered even though ideally it should have been considered an incremental value.

I wondered what the most elegant way to deal with this is? 

I tried to hardcode an exception using an IF statement (e.g. if transitioning from 255 to 0, then 'exception'), but sometimes the numbers increment so fast (e.g. sometimes there will be a transition directly from 240 to 0, rather than 255 to 0) that it doesn't work/may not be the most elegant solution to such a problem.

Can anyone offer some suggestions on this? Thank you.

swipe-dial.maxpat

I have a PS4 controller connected to Max. There is a touch interface which allows me to determine whether I am swiping left or right.
Swiping results in an integer output ranging from 0-255, and swiping left/right leads to a derement/increment (respectively). A particularly quick/long swipe may result in the integer wrapping around before continuing to decrement/increment).

I have attached the method I have implemented so far, which does the job except at the boundary points.

Once the boundary of the integer range is reached, the numbers wrap around, which results in the wrong bang being triggered, for a single instance. e.g. a transition from 255 to 0 results in the left bang being triggered even though ideally it should have been considered an incremental value.

I wondered what the most elegant way to deal with this is? 
I tried to hardcode an exception using an IF statement (e.g. if transitioning from 255 to 0, then 'exception'), but sometimes the numbers increment so fast (e.g. sometimes there will be a transition directly from 240 to 0, rather than 255 to 0) that it doesn't work/may not be the most elegant solution to such a problem.

I have a PS4 controller connected to Max. There is a touch interface which allows me to determine whether I am swiping left or right.
Swipin

how-best-to-deal-with-the-boundary-points-of-a-circular-number-sequence

" I wondered what the most elegant way to deal with this is? "

it should be enough to split the range, and disallow any transition from 235 - 255 to 0- 20 to be analyzed as a negative delta.

how big the range you exclude actually should be ... mostly depends on the data the controller typically produces ... i suppose that it is impossible to create an immediate change from 200 to 45 (upwards, that is) with any type of knob, dial or whatever.

" I wondered what the most elegant way to deal with this is?  "

how big the range you exclude actually should be ...  mostly depends on the data the controller typically produces ... i suppose that it is impossible to create an immediate change from 200 to 45 (upwards, that is) with any type of knob, dial or whatever.

What about calculating the delta (difference of subsequent values) and ignoring those that are unusually large (255) or small (-255)?

I think I've also created a parallel stream of values with an offset added. Then when the original values approach the break point, you can ignore them and look at the offset stream instead.

I think I've also created a parallel stream of values with an  offset added. Then when the original values approach the break point, you can ignore them and look at the offset stream instead.

Yes in the end what I did was the following:

- determine whether there was a jump across the 0/255 boundary and in which direction

- if there was a jump then calculate a 'corrected' delta e.g. for a jump from 5 to 250, the delta is - 11.

- if there isn't a jump then just use the normal delta

- added a slide object at the output to smooth things out and get rid of occasional unwanted sudden spikes.

Yes in the end what I did was the following:
- determine whether there was a jump across the 0/255 boundary and in which direction
- if there was a jump then calculate a 'corrected' delta e.g. for a jump from 5 to 250, the delta is - 11.
- if there isn't a jump then just use the normal delta
- added a slide object at the output to smooth things out and get rid of occasional unwanted sudden spikes. 

Thanks, what I did in the end was the following:

- Determine if there has been a jump across the 0/255 boundary. If so, then calculate a corrected delta (e.g. for a jump from 5 to 250 the delta is -11 rather than 245)

- If there was no jump then just use the normal delta

- Use a slide function at the end to smooth the output out a bit

Thanks, what I did in the end was the following:
- Determine if there has been a jump across the 0/255 boundary. If so, then calculate a corrected delta (e.g. for a jump from 5 to 250 the delta is -11 rather than 245)
- If there was no jump then just use the normal delta
- Use a slide function at the end to smooth the output out a bit

How best to deal with the boundary points of a circular number sequence?