This is a limited unicode-histogram for use in a Google Spreadsheet.

SparkHist.js

function sparkHist(inputRange){
  //this is what we have to work with from the unicode block elements
  //▀▁▂▃▄▅▆▇█▉▊▋▌▍▎▏▐░▒▓▔▕▖▗▘▙▚▛▜▝▞▟
  //▁▂▃▄▅▆▇█ is the useful bit
  var bars=["░","▁","▂","▃","▄","▅","▆","▇","█"];

  //var inputRange = [10,10,8,10,10,10,10,10,1,1,,8,5,10,'[email protected]',,,5,10,7,10,10,1,1,8,7,,7,8,9,9,4,10,,,5,,10,8,8,5,8,10,8,,8,2,10,8,8,8,10,10,8,10,10,5,10,10,5,9,10,8,8,5,,5,,,,,,,,,,10,10,10,9,1,4,8,5,7,,,,8,5,8,,,];
  var filtered = filterNonNum(inputRange[0]);//Logger.log("filtered").log(filtered);
  var binned   = bin(filtered);              //Logger.log(binned);
  var vals     = valueArray(binned);         //Logger.log(vals);
  var most     = Math.max.apply(null, vals); //Logger.log(most);
  var least    = Math.min.apply(null, vals); //Logger.log(least);
  var range    = most-least;                 //Logger.log("**");
  var stepSize = (bars.length-1)/range;      //Logger.log(stepSize);

  var sparkBar=[];
  for (i in binned){
    sparkBar.push(bars[Math.floor(binned[i]*stepSize)]);
  }
  var hist = sparkBar.join("");
  //Logger.log(hist);

  return hist;
}

function filterNonNum(array){
  //tried this with a map, but it puts nulls in when it should skip
  var newArray = [];
  for(i in array){
  var x = array[i];
    if(x !== ""){
      var res = Number(x);
      if (isNaN(res) == false){
        newArray.push(res);
      }
    }
  }
  return newArray;
}


function bin(values){
  var bins = {1:0,
              2:0,
              3:0,
              4:0,
              5:0,
              6:0,
              7:0,
              8:0,
              9:0,
             10:0};
  for(i in values){
    var v = values[i];
    bins[v]++;
  }
  return bins;
}

function valueArray(dataObject){
  var dataArray = new Array;
  for(var o in dataObject) {
      dataArray.push(dataObject[o]);
  }
  return dataArray;
}

It takes a row of values between 1 and 10, bins them and then uses unicode blocks to represent them.

▁░░░▃░▁▅▁█

█▄▁▁▂▃▇▆▂░
▁▄▁▃▃▂▂█▂░
█░░░░░░░░░
█░▃▂█▁▅▁░▁
█▁▂░▁░░░░░
▂░░░░▄▆█▅░
█░▁░░▁▁░░░
█▁▂▁▁░▁▁░░
█░░░░░░░░░
█░░░░░░░░░
█▂▂▁▁░▁░░░
█░░░░░░░░░
▄░▁▁▃░█▅▁▁
▆▂▁░▃▁▄▇█▁
█▁░░▁░▂▁░░
█▅▄▆▁▂▅░▁▁
▇▆█▅▅▃▄▅▂░
█░░░░░░░░░
▂█▃▂▂▂▄█▂

TODO:

• Make it take arbitrary values
• Make it have a controllable number of bins
• Look into how spark tweets are written