Low-variance shifted data about simple-statistics

[plato-cambrian]

here's what I wrote for my use case:

function findShiftedVariance(data){
  if(data.length === 0){ 
    throw new Error('Calculating variance of empty array')
  }

  var n;
  var K;
  var variance;
  var sum;
  var sumSqr;

  K = data[0];
  sum = 0;
  sumSqr = 0;
  for(n=0; n<data.length; n++){
    var x = data[n];
    sum = sum + (x - K);
    sumSqr = sumSqr + ((x - K) * (x - K));
  }
  variance = (sumSqr - ( (sum*sum)/n ) ) / n;
  return variance;
}

[plato-cambrian]

hmm re-reading I note that I am dividing by n in the last step because I have a complete data set; but if the data set is a sample of a larger data set you should divide by n-1. how annoying :D

[tmcw]

Happy to accept a PR implementing this feature, but reading more about these algorithms doesn't uncover any that are sufficiently canonical.