D3: What is a Bisector?

前端 未结 1 1917
天涯浪人
天涯浪人 2020-12-23 04:39

I\'m looking into making charts with D3, and stumbled upon the d3.bisector. However, I don\'t understand what it is or does from the documentation.

Almost all exampl

相关标签:
1条回答
  • 2020-12-23 05:44

    The underlying idea behind bisect is this:

    Consider the array that you mention - var data = [3, 6, 2, 7, 5, 4, 8]

    You want to insert a new value let's say 3.5 into data array and want to know how would that 'partition' it. In other words, you want to know what would be the index of 3.5 if it were inserted when data array is sorted.

       var data = [3, 6, 2, 7, 5, 4, 8]
    
       //Sorted data
    
      [2, 3, 4, 5, 6, 7, 8]
    
      //You want to insert 3.5
    
    
      The sorted array after insertion of 3.5 should look something like:
    
      [2, 3, 3.5, 4, 5, 6, 7, 8]
    
    
      So the index of 3.5 in sorted data array is "2".
    

    There are situations where you would want to know how the insertion of that element 'bisects' or 'divides' an array. In that case, you would want to first sort that array and do what we call a Binary Search to find out the correct position for insertion of that element.

    bisectLeft and bisectRight take care to clarify the anomaly in a situation where you want to enter an element that already exists in the array. Let's say you want to enter another 3 into the array. There are two situations:

       3* -> The new element to be entered
    
    
       [2, 3*, 3, 4, 5, 6, 7, 8] -> entered at "1" (array is still sorted)
    
    
       [2, 3, 3*, 4, 5, 6, 7, 8] -> entered at "2" (array is still sorted)
    

    So depending upon how we take care of this ambiguity, we can enter that element to the 'left' or 'right' of the already existing element. From the docs (Mark the emphasis):

    The returned insertion point i partitions the array into two halves so that all v < x for v in array.slice(lo, i) for the left side and all v >= x for v in array.slice(i, hi) for the right side.

    In bisectLeft we get 1 as the suitable index, all the duplicate entries will be on the right of that index and the situation is exactly the opposite in bisecRight.

    Now that you know how bisectLeft and bisectRight work, the bisector just allows us to define a custom comparator or accessor function to parition the values or make sense of < and > on objects as well.

    So this piece of code:

      var bisect = d3.bisector(function(d) { return d.date; }).right;
    
      var bisect = d3.bisector(function(a, b) { return a.date - b.date; }).right;
    

    Just specifies to use the bisectRight option and return a suitable index for the insertion of an element assuming the array is sorted(in ascending order).

    So if I were to build up on your example and assuming a bisector named bisect. And you did:

     bisect(data, 3); //it would return 2.
    

    I hope it clarifies things and gets you started in the right direction.

    0 讨论(0)
提交回复
热议问题