# Test script used to verify CPU status of the test PC # Recommended to run this script before executing any test suites # ################################################################################ ################################################################################ import sys sys.path.append("../lib") from commonlib import * from TestController import * import xml.etree.ElementTree as ET import re import time #Custom Test libraries cLib = CommonLib() testcontrol = TestController() ################################################################################ # Parse the Command Line Arguments ################################################################################ import argparse def fnTestModule_1(): cLib.logger( "----------------------------------------------------------------------" ) cLib.logger( "Module 1: Basic Functionality - Frequency Shift & Negative Load " ) cLib.logger(
With the initial Berkeley dataset, we've got some backbone traffic in the data. As a result, only a few kinds of 'packets' make up most of our data. Using just IP & Port, roughly 90% (or more) of the traffic is traffic on one of 6 connections. (according to raw packet count) My instinct is generally no, but it might be useful to make KNN's output more sensible. Possibly. I don't know, this may just be an expected result. I'm curious if some of the CRAWDAD data will have similar proprties. I assume it will, if backbone traffic is included. """ # Making a ML system and passing to TestController after this point. kns = KNNSys.KNearestSystem() knc = KNCSys.KNClassifierSystem() controller = TestController.TestController(data) # Working out what the constructor looks like. #knnResults = controller.run(kns) # Data formatting notes # The various scikit-learn libraries expect an 'array like' object. # Fortunately, they can correctly convert a list. So, a list-of-lists is a # valid object to pass them. We may want to convert to array's at some point # for memory reasons, but that can be done hear. Below this, I'll probably # just pass inputs through array() as necessary.