In data science we have "Manhattan Distance" as a slang term for Level 1 Distance and "Euclidean Distance" as a slang term for Level 2 Distance. Is there an accepted term for linear distance in memory of cells in different rows in a matrix? That is, given an 8x8 matrix, the "linear distance" from [3,2] to [4,5] is: (4-3) * 8 + (5-2)

This is the distance in memory addressing. "Level 0 Distance" doesn't really work as a technical term, because computer memory layout is not a native concept in mathematics. Is there a standard term for this? It reminds me of the old flyback transformers in cathode ray tubes, so "flyback distance" makes sense as a slang term.

  • 1
    $\begingroup$ This doesn't strike me as common enough to have a standard name, though I've also never heard "Level i Distance" before, so maybe I'm just coming from the wrong field. I'd suggest also "reading distance" as being evocative of the meaning. $\endgroup$ – Ben Reiniger Jan 29 at 0:07

Actually this is the Manhattan / $L_1$ / Norm 1 distance, if you just multiply the row by $8$ (or $N$).

The $L_1$ distance between $[24, 2]$ and $[32, 5]$ is what you are looking for.

| improve this answer | |
  • $\begingroup$ Right, it's the L1 distance of the matrix linearized. But that linearization is not a mathematical concept, it is just an implementation detail. I'm searching for some reason to not just fabricate my own word, "flyback", for this distance. $\endgroup$ – Jack Parsons Dec 30 '19 at 23:25
  • $\begingroup$ I think this is different: AIUI, the OP's distance from [3,5] to [4,2] is 5, not 11. $\endgroup$ – Ben Reiniger Jan 29 at 0:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.