gistfile1.txt

import csv
import numpy
import time,datetime
import modshogun
import cPickle

vars = {}

# operational context
class operation(object):
	# entering message to be printed, assumed to have two string parameters (timestamp and message)
	__enter__message = '[%s] \033[1;34mStarting %s\033[0m'
	# exit message to be printed, assumed to have three string parameters (timestamp, message and time passed)
	__exit__message = '[%s] \033[0;31mFinished %s, took %fs\033[0m'

	def __init__(self, msg):
		self.msg = msg
		self.parameters = {}
	def __enter__(self):
		self.start = time.time()
		print operation.__enter__message % (datetime.datetime.now(),self.msg)
		return self
	def __exit__(self, type, value, traceback):
		for (k,v) in self.parameters.items():
			with open('%s.dat' % k, 'wb') as output_file:
				cPickle.dump(v,output_file)
		print operation.__exit__message % (datetime.datetime.now(), self.msg, time.time()-self.start)
		return True
	def __getattr__(self,name):
		if name in vars.keys():
			return vars[name]
		raise Exception("Parameter %s is not in the context" % name)
	def __setattr__(self,name,value):
		if name in ['msg', 'parameters','start']:
			object.__setattr__(self,name,value)
		else:
			self.parameters[name] = value
			vars[name] = value
	def __str__(self):
		return self.msg

with operation('reading files') as context:
	context.train_data = numpy.loadtxt('train.csv',delimiter=",",skiprows=1)
	context.test_data = numpy.loadtxt('test.csv',delimiter=",",skiprows=1)

with operation('training model') as context:
	context.classifier = modshogun.MulticlassLibLinear()
	context.train_features = modshogun.ByteFeatures(context.train_data)
	context.test_features = modshogun.ByteFeatures(context.test_data)