CSC352 Game of Life in Map-Reduce

From dftwiki3
Revision as of 09:00, 21 April 2017 by Thiebaut (talk | contribs) (→‎Reducer.py)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

--D. Thiebaut (talk) 16:47, 2 April 2017 (EDT)

Reference



Introduction


  • We map-reduce once to go from one generation to the next.
  • The original array (see Data File section below) is transformed into a file where each line is numbered, and each row of the game of life is sandwiched between double quotes.
  • To go from the data file to the first generation:
cat life0.txt | ./mapper.py | ./shuffleSort.py | ./reducer.py  
2	4
2	5
2	6
the output is a horizontal line, on Row 2, spanning Columns 4, 5, and 6.
  • To generate the second generation:
cat life0.txt | ./mapper.py | ./shuffleSort.py | ./reducer.py | ./mapper.py | ./shuffleSort.py | ./reducer.py
1	5
2	5
3	5
  • To generate the third generation:
cat life0.txt | ./mapper.py | ./shuffleSort.py | ./reducer.py | ./mapper.py | ./shuffleSort.py | ./reducer.py \
   | ./mapper.py | ./shuffleSort.py | ./reducer.py
2	4
2	5
2	6

Data File


The file is called life0.txt. 

0       "           "
1       "     *     "
2       "     *     "
3       "     *     "
4       "           "


Mapper.py


#!/usr/bin/env python
# mapper.py 
# mapper for Conway's 2D Game of Life.
# The input data has two possible formats:
# 1) The first first generation defined by 
# a python-type collection of quoted strings.
# We refer to it  as "raw-life"
#
#0       "           "
#1       "     *     "
#2       "     *     "
#3       "     *     "
#4       "           "
#
# 2) the output comes from a previous map-reduce stage,
# in which case the reducer outputs the coordinates of 
# live cells only:
#
# 2  4
# 2  5
# 2  6
# 
# Above, 3 cells are live, at coordinates (2,4), (2,5), and (2,6)
#
# For every live cell that it finds, the mapper sends 9 tupples:
# 001005 alive
# 000004 OneMoreNeighbor
# 000005 OneMoreNeighbor
# 000006 OneMoreNeighbor
# 001004 OneMoreNeighbor
# 001006 OneMoreNeighbor
# 002004 OneMoreNeighbor
# 002005 OneMoreNeighbor
# 002006 OneMoreNeighbor 
#
# The first string contains the 0-padded coordinates of live cells and
# neighbor cells that have this live cell as a neighbor.
# 

from __future__ import print_function
import sys

# input comes from STDIN (standard input)
for line in sys.stdin:

    # remove leading and trailing whitespace
    line = line.strip()
    #print( "line =", line )

    # split the line into words
    words = line.split( None, 1 )

    #print( "words = ", words )
    isRawLife = line.find( "\"" ) != -1

    # are we reading the original array of cells from a file?
    if isRawLife:
       lineNo = int( words[0] )
       lineWidth = len( words[1] )-2 # for the double quotes
       lifeLine = words[1].replace( "\"", "" )
       #print( "(%d, >%s<)" % ( lineNo, lifeLine ) )
       for j in range( len( lifeLine ) ):
           if lifeLine[j] != ' ':
              print( "%03d%03d\t%s" % ( lineNo, j, "alive" ) )
              for row in range( lineNo-1, lineNo+1+1 ):
                  for col in range( j-1, j+1+1 ):
                      if row < 0: continue
                      if col < 0: continue
                      if col > lineWidth: col = 0
                      if row == lineNo and col == j: continue
                      print( "%03d%03d\t%s" % ( row, col, "OneMoreNeighbor" ) )
     
    # or are we reading tuples output by the reducer?                    
    else:
       lineNo = int( words[0] )
       j      = int( words[1] )
       print( "%03d%03d\t%s" % ( lineNo, j, "alive" ) )
       for row in range( lineNo-1, lineNo+1+1 ):
           for col in range( j-1, j+1+1 ):
               if row < 0: continue
               if col < 0: continue
               if row == lineNo and col == j: continue
               print( "%03d%03d\t%s" % ( row, col, "OneMoreNeighbor" ) )


Mapper on Original Life File


cat life0.txt | ./mapper.py 
001005	alive
000004	OneMoreNeighbor
000005	OneMoreNeighbor
000006	OneMoreNeighbor
001004	OneMoreNeighbor
001006	OneMoreNeighbor
002004	OneMoreNeighbor
002005	OneMoreNeighbor
002006	OneMoreNeighbor
002005	alive
001004	OneMoreNeighbor
001005	OneMoreNeighbor
001006	OneMoreNeighbor
002004	OneMoreNeighbor
002006	OneMoreNeighbor
003004	OneMoreNeighbor
003005	OneMoreNeighbor
003006	OneMoreNeighbor
003005	alive
002004	OneMoreNeighbor
002005	OneMoreNeighbor
002006	OneMoreNeighbor
003004	OneMoreNeighbor
003006	OneMoreNeighbor
004004	OneMoreNeighbor
004005	OneMoreNeighbor
004006	OneMoreNeighbor


Mapper on Output of Reduce


cat life0.txt | ./mapper.py | ./shuffleSort.py | ./reducer.py | ./mapper.py
002004	alive
001003	OneMoreNeighbor
001004	OneMoreNeighbor
001005	OneMoreNeighbor
002003	OneMoreNeighbor
002005	OneMoreNeighbor
003003	OneMoreNeighbor
003004	OneMoreNeighbor
003005	OneMoreNeighbor
002005	alive
001004	OneMoreNeighbor
001005	OneMoreNeighbor
001006	OneMoreNeighbor
002004	OneMoreNeighbor
002006	OneMoreNeighbor
003004	OneMoreNeighbor
003005	OneMoreNeighbor
003006	OneMoreNeighbor
002006	alive
001005	OneMoreNeighbor
001006	OneMoreNeighbor
001007	OneMoreNeighbor
002005	OneMoreNeighbor
002007	OneMoreNeighbor
003005	OneMoreNeighbor
003006	OneMoreNeighbor
003007	OneMoreNeighbor


ShufflerSort.py


#!/usr/bin/env python
from __future__ import print_function
import sys

sortNumeric = False
if len( sys.argv ) > 1 and sys.argv[1] == "-n":
   sortNumeric = True

L = []

# input comes from STDIN
for line in sys.stdin:
      # remove leading and trailing whitespace
      line = line.strip()

      word, count = line.split('\t', 1)
      if sortNumeric:
          word = int( word )

      L.append( (word, count) )


L.sort()

for word, count in L:
    print( '%s\t%s' % (word, count) )


Output of ShuffleSort


cat life0.txt | ./mapper.py | ./shuffleSort.py 
000004	OneMoreNeighbor
000005	OneMoreNeighbor
000006	OneMoreNeighbor
001004	OneMoreNeighbor
001004	OneMoreNeighbor
001005	OneMoreNeighbor
001005	alive
001006	OneMoreNeighbor
001006	OneMoreNeighbor
002004	OneMoreNeighbor
002004	OneMoreNeighbor
002004	OneMoreNeighbor
002005	OneMoreNeighbor
002005	OneMoreNeighbor
002005	alive
002006	OneMoreNeighbor
002006	OneMoreNeighbor
002006	OneMoreNeighbor
003004	OneMoreNeighbor
003004	OneMoreNeighbor
003005	OneMoreNeighbor
003005	alive
003006	OneMoreNeighbor
003006	OneMoreNeighbor
004004	OneMoreNeighbor
004005	OneMoreNeighbor
004006	OneMoreNeighbor


Reducer.py


#!/usr/bin/env python
# reducer.py
# D. Thiebaut
# Reducer stage for the game of life.
# The input typically looks like this:
# 
# 001005 alive
# 000004 OneMoreNeighbor
# 000005 OneMoreNeighbor
# 000006 OneMoreNeighbor
# 001004 OneMoreNeighbor
# 001006 OneMoreNeighbor
# 002004 OneMoreNeighbor
# 002005 OneMoreNeighbor
# 002006 OneMoreNeighbor 
#
# where the first word is a string of 6 0-padded digits, the first 3 
# representing the row of a cell, and the last 3 its column.  The
# second word indicates whether the cell is actually a live cell in 
# the current generation, or whether its status is unknown, but there
# is a cell in the previous generation that is alive and is its 
# neighbor.
#
# Since all the lines fed to the reducer are sorted, then lines 
# with the same keys, representing the same cell, are together.
# 
# 000004   OneMoreNeighbor
# 000005   OneMoreNeighbor
# 000006   OneMoreNeighbor
# 001004   OneMoreNeighbor
# 001004   OneMoreNeighbor
# 001005   alive
# 001005   OneMoreNeighbor
# 001006   OneMoreNeighbor
# 001006   OneMoreNeighbor
# 002004   OneMoreNeighbor
# 002004   OneMoreNeighbor
# 002004   OneMoreNeighbor
# 002005   alive
# 002005   OneMoreNeighbor
# 002005   OneMoreNeighbor
# 002006   OneMoreNeighbor
# 002006   OneMoreNeighbor
# 002006   OneMoreNeighbor
# 003004   OneMoreNeighbor
# 003004   OneMoreNeighbor
# 003005   alive
# 003005   OneMoreNeighbor
# 003006   OneMoreNeighbor
# 003006   OneMoreNeighbor
# 004004   OneMoreNeighbor
# 004005   OneMoreNeighbor
# 004006   OneMoreNeighbor
#
# It is then easy to divide the input lines into blocks with the same key,
# and count the number of neighbors, and know if the cell is alive or dead.
# If the cell should be alive in the next generation, then the reducer
# outputs its coordinates (row, column) as two numbers separated by a tab.
# For example:
#
# 2  4
# 2  5
# 2  6
# 
#

from __future__ import print_function
from operator import itemgetter

import sys

lastCell = None
isAlive = False
cell = None
noNeighbors = 0

# input comes from STDIN
for line in sys.stdin:
      # remove leading and trailing whitespace
      line = line.strip()
      
      # parse the input we got from mapper.py
      cell, status = line.split('\t', 1)
      
      if cell != lastCell:
         # decide of fate of last cell
         if isAlive and 2 <= noNeighbors <= 3:
            # continue living
            print( "%d\t%d" % ( int( lastCell )/1000, int( lastCell )%1000 ) )
         if not isAlive and noNeighbors == 3:
            # be born 
            print( "%d\t%d" % ( int( lastCell )/1000, int( lastCell )%1000 ) )

         noNeighbors = 0
         isAlive = False
 
      # parse status of new cell
      if status.find( "alive" ) != -1:
          isAlive = True
      else:
          noNeighbors += 1

      lastCell = cell


# do not forget to output the last cell if needed!
if isAlive and 2 <= noNeighbors <= 3:
   # continue living
   print( "%d\t%d" % ( int( lastCell )/1000, int( lastCell )%1000 ) )
if not isAlive and noNeighbors == 3:
   # be born
   print( "%d\t%d" % ( int( lastCell )/1000, int( lastCell )%1000 ) )


Output of Reducer


cat life0.txt | ./mapper.py | ./shuffleSort.py | ./reducer.py  
2	4
2	5
2	6





















Retrieved from "http://dominiquethiebaut.com/dftwiki3/index.php?title=CSC352_Game_of_Life_in_Map-Reduce&oldid=30909"