Upload project Structure

.gitignore

@@ -10,6 +10,12 @@ local.properties
 .loadpath
 .recommenders
 
+#For omitting Data
+data/
+
+#For csv file
+*.csv
+
 # Eclipse Core
 .project
 

src/ManipulateData.py

@@ -0,0 +1,110 @@
+'''
+Created on Sep 24, 2016
+
+@author: sarker
+'''
+import csv
+import numpy as np
+
+def getDataMatrixFromCSV(fileName):
+    dataMatrix = np.array(list(csv.reader(open(fileName, "r+"), delimiter=',')))
+
+    return dataMatrix
+
+def shuffleData(dataMatrix):
+    dataMatrix = np.random.permutation(dataMatrix)
+    return dataMatrix
+
+def convertDatatoFloat(dataMatrix):
+    dataMatrixAsfloat = [ [ np.float128(eachVal) for eachVal in row ] for row in dataMatrix]
+    return dataMatrixAsfloat   
+
+def convertDatatoZeroOne(dataList):
+    dataAsZeroOne = [] 
+    for eachVal in dataList:
+        if(eachVal == 'N'):
+            dataAsZeroOne.append(0)
+        else:
+            dataAsZeroOne.append(1)
+
+    return dataAsZeroOne 
+
+def getNormalizedData(dataMatrix):
+    normalizedDatas = [] 
+    for dataList in dataMatrix:
+        _mean = sum(dataList) / float(len(dataList))
+        _max = max(dataList)
+        _min = min(dataList)
+
+        normalizedData = []
+
+        for eachVal in dataList:     
+            normalizedData.append((eachVal - _mean) / (_max - _min))
+        normalizedDatas.append(normalizedData)
+
+    return np.array(normalizedDatas)
+
+def getVectorizedClassValues(classes):
+    OutputLayerNoOfNeuron = len(set(classes)) 
+
+    outputVector = np.zeros((len(classes), OutputLayerNoOfNeuron))
+
+    for i in range(0, len(classes)):
+        if(classes[i] == 1):
+            outputVector[i] = [1, 0, 0]
+        elif(classes[i] == 2):
+            outputVector[i] = [0, 1, 0]
+        elif(classes[i] == 3):
+            outputVector[i] = [0, 0, 1]
+
+    return outputVector
+
+
+def getVectorizedClassValuesFromYesNo(classes):
+    OutputLayerNoOfNeuron = len(set(classes)) 
+
+    outputVector = np.zeros((len(classes), OutputLayerNoOfNeuron))
+
+    for i in range(0, len(classes)):
+        if(classes[i] == 'N'):
+            outputVector[i] = [0, 1]
+        elif(classes[i] == 'R'):
+            outputVector[i] = [1, 0]
+
+    return outputVector
+
+def zeroOneClassCounter(dataList):
+    classZero = 0
+    classOne = 0
+    for i in dataList:
+        if(i == 0):
+            classZero += 1
+        elif(i == 1):
+            classOne += 1
+    return classZero, classOne
+
+def splitintoNFold(l, n):
+    n = max(1, n)
+    return (l[i:i + n] for i in range(0, len(l), n))
+
+def splitTrainAndTestData(data, ratio):
+    trainData = data[:int(len(data) * ratio)]
+    testData = data[int(len(data) * ratio):]
+    return trainData, testData
+
+def splitTrainValidateAndTestData(inputData, outputVector, ratio1, ratio2, ratio3):
+    dataLength = len(inputData)
+    if(ratio1 + ratio2 + ratio3 != 1):
+        print('Sum of ratio must be 1')
+        return 'Sum of ratio must be 1'
+    trainInputData = inputData[ :int(dataLength * ratio1) ]
+    trainOutputVector = outputVector[ :int(dataLength * ratio1) ]
+
+    validationInputData = inputData[ int(dataLength * ratio1) : int(dataLength * (ratio1 + ratio2)) ]
+    validationOutputVector = outputVector[ int(dataLength * ratio1) : int(dataLength * (ratio1 + ratio2)) ]
+
+    testInputData = inputData[ int(dataLength * (ratio1 + ratio2)) :  ]
+    testOutputVector = outputVector[ int(dataLength * (ratio1 + ratio2)) :  ]
+
+    return trainInputData, trainOutputVector, validationInputData, validationOutputVector, testInputData, testOutputVector
+