Python GAUtility Examples

Example #1

File: optimizeErdosRenyiGraphConnectivity.py Project: rajcscw/echo-state-networks

# Form feature vectors for training data
trainingInputData, trainingOutputData = util.formFeatureVectors(trainingData)
actualOutputData = minMax.inverse_transform(testingData)[:, 0]

# Initial seed
initialSeedForValidation = trainingData[-1]
networkSize = 500
populationSize = 10
noOfBest = int(populationSize / 2)
noOfGenerations = 10
predictedOutputData, bestPopulation = utilityGA.tuneTrainPredictConnectivityGA(
    trainingInputData=trainingInputData,
    trainingOutputData=trainingOutputData,
    validationOutputData=validationData,
    initialInputSeedForValidation=initialSeedForValidation,
    horizon=nTesting,
    noOfBest=noOfBest,
    resTopology=utilityGA.Topology.ErdosRenyi,
    size=networkSize,
    popSize=populationSize,
    maxGeneration=noOfGenerations,
)

predictedOutputData = minMax.inverse_transform(predictedOutputData)

# Plotting of the prediction output and error
outputFolderName = "Outputs/Erdos_Renyi_Outputs" + datetime.now().strftime("%Y_%m_%d_%H_%M_%S")
os.mkdir(outputFolderName)
outplot = outputPlot.OutputPlot(
    outputFolderName + "/Prediction.html",
    "Mackey-Glass Time Series - GA Optimization)",
    "Prediction on Testing data",

Example #2

File: plotForSmallWorld.py Project: nivedita/echo-state-networks

from reservoir import GAUtility as utilityGA
from plotting import ScatterPlot as plot
import numpy as np
import os
from datetime import datetime

# File name for
folderName = "Outputs/GAResults_Small_World_Graphs_" + datetime.now().strftime(
    "%Y_%m_%d_%H_%M_%S")
os.mkdir(folderName)
# Load the best population from the file
bestPopulation = utilityGA.loadBestPopulation("population.pkl")

# Iterate over all the elements and get the network properties
networkSize = 500
meanDegreeList = []
betaList = []
errorList = []
averageDegreeList = []
averagePathLengthList = []
averageDiameterList = []
averageClusteringCoefficientList = []
for item in bestPopulation:
    # Fitness
    meanDegree = item[0][0, 0]
    beta = item[0][0, 1]
    error = item[1]

    # Network properties
    averageDegree = item[2]['averageDegree']
    averagePathLength = item[2]['averagePathLength']

Example #3

File: testGAOptimizer.py Project: rajcscw/echo-state-networks

# Split the data into training, validation and testing
trainingData, validationData, testingData = util.splitData(data, 0.5, 0.25, 0.25)
nValidation = validationData.shape[0]
nTesting = testingData.shape[0]

# Form feature vectors for training data
trainingInputData, trainingOutputData = util.formFeatureVectors(trainingData)
actualOutputData = minMax.inverse_transform(testingData)[:,0]

# Initial seed
initialSeedForValidation = trainingData[-1]

predictedOutputData = utilGA.tuneTrainPredictGA(trainingInputData=trainingInputData,
                                            trainingOutputData=trainingOutputData,
                                            validationOutputData=validationData,
                                            initialInputSeedForValidation=initialSeedForValidation,
                                            testingData=actualOutputData
                                            )


predictedOutputData = minMax.inverse_transform(predictedOutputData)

#Plotting of the prediction output and error
outputFolderName = "Outputs/Outputs" + datetime.now().strftime("%Y_%m_%d_%H_%M_%S")
os.mkdir(outputFolderName)
outplot = outputPlot.OutputPlot(outputFolderName + "/Prediction.html", "Mackey-Glass Time Series - GA Optimization", "Predicted vs Actual", "Time", "Output")
outplot.setXSeries(np.arange(1, nTesting + 1))
outplot.setYSeries('Actual Output', actualOutputData)
outplot.setYSeries('Predicted Output', predictedOutputData)
outplot.createOutput()

Example #4

File: optimizeRandomGraphConnectivity.py Project: nivedita/echo-state-networks

# Form feature vectors for training data
trainingInputData, trainingOutputData = util.formFeatureVectors(trainingData)
actualOutputData = minMax.inverse_transform(testingData)[:, 0]

# Initial seed
initialSeedForValidation = trainingData[-1]
networkSize = 500
populationSize = 200
noOfBest = int(populationSize / 2)
noOfGenerations = 100
predictedOutputData, bestPopulation = utilityGA.tuneTrainPredictConnectivityGA(
    trainingInputData=trainingInputData,
    trainingOutputData=trainingOutputData,
    validationOutputData=validationData,
    initialInputSeedForValidation=initialSeedForValidation,
    horizon=nTesting,
    noOfBest=noOfBest,
    resTopology=utilityGA.Topology.Random,
    size=networkSize,
    popSize=populationSize,
    maxGeneration=noOfGenerations)

predictedOutputData = minMax.inverse_transform(predictedOutputData)

#Plotting of the prediction output and error
outputFolderName = "Outputs/Random_Graph_Outputs" + datetime.now().strftime(
    "%Y_%m_%d_%H_%M_%S")
os.mkdir(outputFolderName)
outplot = outputPlot.OutputPlot(outputFolderName + "/Prediction.html",
                                "Mackey-Glass Time Series - GA Optimization)",
                                "Prediction on Testing data", "Time", "Output")

Example #5

# Split the data into training, validation and testing
trainingData, validationData, testingData = util.splitData(
    data, 0.5, 0.25, 0.25)
nValidation = validationData.shape[0]
nTesting = testingData.shape[0]

# Form feature vectors for training data
trainingInputData, trainingOutputData = util.formFeatureVectors(trainingData)
actualOutputData = minMax.inverse_transform(testingData)[:, 0]

# Initial seed
initialSeedForValidation = trainingData[-1]

predictedOutputData = utilGA.tuneTrainPredictGA(
    trainingInputData=trainingInputData,
    trainingOutputData=trainingOutputData,
    validationOutputData=validationData,
    initialInputSeedForValidation=initialSeedForValidation,
    testingData=actualOutputData)

predictedOutputData = minMax.inverse_transform(predictedOutputData)

#Plotting of the prediction output and error
outputFolderName = "Outputs/Outputs" + datetime.now().strftime(
    "%Y_%m_%d_%H_%M_%S")
os.mkdir(outputFolderName)
outplot = outputPlot.OutputPlot(outputFolderName + "/Prediction.html",
                                "Mackey-Glass Time Series - GA Optimization",
                                "Predicted vs Actual", "Time", "Output")
outplot.setXSeries(np.arange(1, nTesting + 1))
outplot.setYSeries('Actual Output', actualOutputData)
outplot.setYSeries('Predicted Output', predictedOutputData)

Example #6

from reservoir import GAUtility as utilityGA
from plotting import ScatterPlot as plot
import numpy as np
import os
from datetime import datetime

# File name for
folderName = "Outputs/GAResults_Combined_" + datetime.now().strftime(
    "%Y_%m_%d_%H_%M_%S")
os.mkdir(folderName)

# Load the best population from the file
bestPopulationErdos = utilityGA.loadBestPopulation("erdos.pkl")
bestPopulationScaleFree = utilityGA.loadBestPopulation("scalefree.pkl")
bestPopulationSmallWorld = utilityGA.loadBestPopulation("smallworld.pkl")
bestPopulationList = [
    bestPopulationErdos, bestPopulationScaleFree, bestPopulationSmallWorld
]

# Iterate over all the elements and get the network properties
networkSize = 500
errorList = []
averageDegreeList = []
averagePathLengthList = []
averageDiameterList = []
averageClusteringCoefficientList = []

for bestPopulation in bestPopulationList:

    for item in bestPopulation:
        # Fitness

Example #7

File: plotForRandom.py Project: nivedita/echo-state-networks

from reservoir import GAUtility as utilityGA
from plotting import ScatterPlot as plot
import numpy as np
import os
from datetime import datetime

# File name for
folderName = "Outputs/GAResults_Random_Graphs_" + datetime.now().strftime(
    "%Y_%m_%d_%H_%M_%S")
os.mkdir(folderName)
# Load the best population from the file
bestPopulation = utilityGA.loadBestPopulation("population.pkl")

# Iterate over all the elements and get the network properties
networkSize = 500
connectivityList = []
errorList = []
averageDegreeList = []
averagePathLengthList = []
averageDiameterList = []
averageClusteringCoefficientList = []
for item in bestPopulation:
    # Fitness
    connectivity = item[0][0, 0]
    error = item[1]

    # Network properties
    averageDegree = item[2]['averageDegree']
    averagePathLength = item[2]['averagePathLength']
    averageDiameter = item[2]['averageDiameter']
    averageClusteringCoefficient = item[2]['averageClusteringCoefficient'] * 500

Example #8

File: optimizeSmallWorldGraphs.py Project: nivedita/echo-state-networks

# Form feature vectors for training data
trainingInputData, trainingOutputData = util.formFeatureVectors(trainingData)
actualOutputData = minMax.inverse_transform(testingData)[:, 0]

# Initial seed
initialSeedForValidation = trainingData[-1]
networkSize = 500
populationSize = 10
noOfBest = int(populationSize / 2)
noOfGenerations = 10
predictedOutputData, bestPopulation = utilityGA.tuneTrainPredictConnectivityGA(
    trainingInputData=trainingInputData,
    trainingOutputData=trainingOutputData,
    validationOutputData=validationData,
    initialInputSeedForValidation=initialSeedForValidation,
    horizon=nTesting,
    noOfBest=noOfBest,
    resTopology=utilityGA.Topology.SmallWorldGraphs,
    size=networkSize,
    popSize=populationSize,
    maxGeneration=noOfGenerations)

predictedOutputData = minMax.inverse_transform(predictedOutputData)

#Plotting of the prediction output and error
outputFolderName = "Outputs/Small_World_Graphs" + datetime.now().strftime(
    "%Y_%m_%d_%H_%M_%S")
os.mkdir(outputFolderName)
outplot = outputPlot.OutputPlot(outputFolderName + "/Prediction.html",
                                "Mackey-Glass Time Series - GA Optimization)",
                                "Prediction on Testing data", "Time", "Output")

Example #9

File: plotForSmallWorld.py Project: rajcscw/echo-state-networks

from reservoir import GAUtility as utilityGA
from plotting import ScatterPlot as plot
import numpy as np
import os
from datetime import datetime


# File name for
folderName = "Outputs/GAResults_Small_World_Graphs_" + datetime.now().strftime("%Y_%m_%d_%H_%M_%S")
os.mkdir(folderName)
# Load the best population from the file
bestPopulation = utilityGA.loadBestPopulation("population.pkl")


# Iterate over all the elements and get the network properties
networkSize = 500
meanDegreeList = []
betaList = []
errorList = []
averageDegreeList = []
averagePathLengthList = []
averageDiameterList = []
averageClusteringCoefficientList = []
for item in bestPopulation:
    # Fitness
    meanDegree = item[0][0,0]
    beta = item[0][0,1]
    error = item[1]

    # Network properties
    averageDegree = item[2]['averageDegree']

Example #10

File: plotCombined.py Project: rajcscw/echo-state-networks

from reservoir import GAUtility as utilityGA
from plotting import ScatterPlot as plot
import numpy as np
import os
from datetime import datetime


# File name for
folderName = "Outputs/GAResults_Combined_" + datetime.now().strftime("%Y_%m_%d_%H_%M_%S")
os.mkdir(folderName)

# Load the best population from the file
bestPopulationErdos = utilityGA.loadBestPopulation("erdos.pkl")
bestPopulationScaleFree = utilityGA.loadBestPopulation("scalefree.pkl")
bestPopulationSmallWorld = utilityGA.loadBestPopulation("smallworld.pkl")
bestPopulationList = [bestPopulationErdos, bestPopulationScaleFree, bestPopulationSmallWorld]


# Iterate over all the elements and get the network properties
networkSize = 500
errorList = []
averageDegreeList = []
averagePathLengthList = []
averageDiameterList = []
averageClusteringCoefficientList = []

for bestPopulation in bestPopulationList:

    for item in bestPopulation:
        # Fitness
        attachment = item[0][0,0]/networkSize