Parallel Computing for Data Science

4.9. EXAMPLE: K-MEANS CLUSTERING 109

clu s t e r EvalQ ( c l s , kmeans ( x , 2 , 1 , cn trds , sms , ” l c k ” ,

c i n i t=rbind ( c ( 5 , 5 ) , c ( 1 5 , 1 5 ) ) ) )

}

t e s t 1 <− function ( c l s ) {

mg r i ni t ( c l s )

mgrmakevar ( c l s , ”x” , 100 0 0 ,3)

mgrmakevar ( c l s , ” c nt r ds ” , 3 ,3 )

mgrmakevar ( c l s , ”sms” , 3 ,4 )

mgrmakelock ( c l s , ” l c k ” )

x [ , ] <− matrix (rnorm(3 0 0 00 ) , ncol=3)

r i <− sample ( 1 : 1 0 0 0 0 , 3 00 0 )

x [ r i , 1 ] <− x [ r i , 1 ] + 5

r i <− sample ( 1 : 1 0 0 0 0 , 3 00 0 )

x [ r i , 2 ] <− x [ r i , 2 ] + 5

c l u s t e r E x p o r t ( c l s , ”kmeans” )

clu s t e r EvalQ ( c l s , kmeans ( x