I am looking for the distance between the spatial median of a subset of points to all original points in the set preferences to model group decision making. I can build a for loop to return the subsets spatial medians but am having trouble extracting the optimized distances. Any help is greatly appreciated.

n = 5
preferences = RandomVariate[NormalDistribution[5, 2], {n, 2}]
dtotal[f_] := Total[Sqrt[Total[(Transpose[f] - {x, y})^2]]]
sodtotal = FindMinimum[dtotal[preferences], {x}, {y}]
sopt = {x /. sodtotal[[2]], y /. sodtotal[[2]]}
sodist = sodtotal[[1]]
prefsub = Subsets[preferences, {1, n}]
subdtotal = Map[dtotal, Subsets[preferences, {1, n}]]
subopt = {}
For[i = 1, i <= Length[subdtotal], i++, 
 subopt = Append[
   subopt, {x /. FindMinimum[subdtotal[[i]], {x}, {y}][[2]], 
    y /. FindMinimum[subdtotal[[i]], {x}, {y}][[2]]}]]

Creating Lists From Loops How do I find the Euclidean distance between one point and all the points in a list?

  • $\begingroup$ I think it should look something like subsetdist = {} For[i = 1, i <= Length[subdtotal], i++, subsetdist = Append[subsetdist, {FindMinimum[subdtotal[[i]], {x}, {y}, [[1]]]}]] but I am really trying to get away from this For loop. $\endgroup$ Nov 2 '17 at 20:42
  • $\begingroup$ a bit aside but NMinimize works better here instead of FindMinimum $\endgroup$
    – george2079
    Nov 2 '17 at 21:20
  • $\begingroup$ Thanks I will change them. $\endgroup$ Nov 2 '17 at 21:29

You can replace your subopt initialization and the For loop with:

subopt = {x, y} /. FindMinimum[#, {x}, {y}][[2]] & /@ subdtotal

Or using Map explicitly

{x, y} /. Map[FindMinimum[#, {x}, {y}][[2]] &, subdtotal]
  • $\begingroup$ Thank you that Map function replaces the loop very well. I was hoping to use the Nminimize function for the distance from each point in subopt to all points in preferences. $\endgroup$ Nov 2 '17 at 21:25

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.