Question

我有一个numpy数组-

mapPartitionsWithInputSplit(new Function2<InputSplit, Iterator<Tuple2<LongWritable, Text>>, Iterator<Tuple2<String,String>> >(){

    @Override
    public Iterator<Tuple2<String,String>> call(InputSplit arg0,
            Iterator<Tuple2<LongWritable, Text>> dataIterator) throws Exception {
            FileSplit fileSplit = (FileSplit) arg0;
            //Retrieve the file name from the split
            String fileLocation = fileSplit.getPath().toString();
            List<Tuple2<String,String>> retList = new LinkedList<Tuple2<String,String>>();
            while(dataIterator.hasNext())
            {
                String data = dataIterator.next()._2.toString();
                retList.add(new Tuple2<String,String>(fileLocation,data));
            }
            return retList.iterator();
    }   
},true)

，并希望按第1列然后按第0列对其进行排序，同时保留行顺序。我设法按轴1对其进行了排序-

 a = np.array([[1,23],[5,4],[56,12], [7,29], [3,4], [2,15],[2,8], [4,8]])

但第0列也必须在第1列的升序内按升序排列，例如：

 a = a[a[:,1].argsort()]

 array([[ 5,  4],
   [ 3,  4],
   [ 4,  8],
   [ 2,  8],
   [56, 12],
   [ 2, 15],
   [ 1, 23],
   [ 7, 29]])

如何实现？

Answer 1

尝试通过多个键使用sorted：

a = np.array([[1,23],[5,4],[56,12], [7,29], [3,4], [2,15],[2,8], [4,8]])

>>> np.array(sorted(a, key=lambda x: (x[1], x[0]))) # sort col 1 then col 0.
array([[ 3,  4],
       [ 5,  4],
       [ 2,  8],
       [ 4,  8],
       [56, 12],
       [ 2, 15],
       [ 1, 23],
       [ 7, 29]])

按一列然后按另一列（作为子集）对numpy进行排序，同时保留行顺序

1 个答案: