Command-line flag to tell Spark whether data are row or column-oriented

[magsol]

Command-line flag to tell Spark whether data are row or column-oriented

[MOJTABAFA]

@magsol
with following code we can determine whether our data is row-based or column-based :

  parser.add_argument('--coulumnwise', '-c', action='store_true', default=False,
        help = 'Monitor input file for changes')
    args = vars(parser.parse_args())
    if args['coulumnwise'] == True : 
        print ('The data is coulumn-based')
    else :
        print ('The data is Row-wise')

I'm still working on code consistency , for debugging the code I already installed eClipse and set the Pydev and spark there. One thing more interesting is : in debugging the data you can easily see the internal python functions. For instance when we're using the sla.norm or argpartition we can easily see how pyspark and python are calling lower level functions and it may give better understanding about the fact that which part of code should be optimized more.