class Solution(object):
 
 def findmedian(self, lists):
  if not lists or len(lists) == 0:
   return []
 
  n = len(lists)
  if n % 2 == 0:
   a = self.partition(lists, n/2, 0, n-1)
   b = self.partition(lists, n/2-1, 0, n-1)
   mid = (lists[a]+lists[b])/ (2 * 1.0)
   return mid
  else:
   mid = self.partition(lists, n/2, 0, n-1)
   return lists[mid]
 
 def partition(self, lists, k, start, end):
 
  key = lists[start]
  left, right = start, end
 
  while left < right:
   while left < right and lists[right] > key:
    right = right - 1
   lists[left] = lists[right]
 
   while left < right and lists[left] < key:
    left = left + 1
   lists[right] = lists[left]
 
  lists[left] = key
 
  if left == k:
   return left
  elif left > k:
   return self.partition(lists, k, start, left-1)
  else:
   return self.partition(lists, k, left+1, end)
 
 
if __name__ == "__main__":
 
 sol = Solution()
 lists = [2, 5, 4, 9, 3, 6, 8, 7, 1]
 # lists = [1, 2]
 
 
 data = sol.findmedian(lists)
 print("中位數(shù) = %s" % data)

#!/bin/bash

export CURRENT=/home/chunhe.liao/hadoop_streaming/sort

/usr/local/hadoop-2.6.3/bin/hadoop jar /usr/local/hadoop-2.6.3/share/hadoop/tools/lib/hadoop-streaming-2.6.3.jar \
-D stream.map.output.field.separator='\t' \
-D stream.num.map.output.key.fields=3 \
-D mapreduce.job.output.key.comparator.class=org.apache.hadoop.mapreduce.lib.partition.KeyFieldBasedComparator \
-D mapreduce.partition.keycomparator.options=-k3,3nr \  # 按照第三列逆序排列,可以根據(jù)想要的第幾段來選擇。
-input "/user/test/inputdata/datas3/data.txt" \
-output "/user/test/streaming/sorted_20180711" \
-mapper "python mapper.py" \
-reducer "python reducer.py" \
-file "$CURRENT/mapper.py" \
-file "$CURRENT/reducer.py"

# -*- coding: utf-8 -*-
import sys
 
for line in sys.stdin:
 line = line.strip()
 print('{0}'.format(line))

# -*- coding: utf-8 -*-
import sys
 
for line in sys.stdin:
 line = line.strip()
 print("{0}".format(line))