def main():
x = gam.getAdjMat('redwine.csv', False) ## getAdjMatrix1.getmatrix returns a tuple
std = x[0]
mean = x[1]
matrix = x[2]
thetas = np.array([1.10E-01,-1.80E-01,-1.30E-02,-6.59E-03,-9.09E-02,3.40E-02, -6.71E-02, -1.29E-01, -7.13E-02, 2.01E-01, 3.77E-01 ,-7.33E-02])
predictedValuesT = predictValue('redwineTest.csv', thetas, False, std, mean, "predictedTestResult1.csv", False)
print reportError(predictedValuesT, 'redwineTestResult.csv')
array = []
for i in range(60):
alpha = random.uniform(0.0001, 0.01)
gamma = random.uniform(0.00000001, 0.00001)
results = regressionAndTestError('whitewineTraining.csv', 'whiteWineVal.csv',
'whitewineValResult.csv', 'whitewineTest.csv', 'whitewineTestResult.csv', alpha,
0.0001, gamma, False, False, False)
array.append(results)
print "\n\nLoop ", i
np.savetxt("RegressionResults.csv", np.asarray(array), delimiter=",")
def regressionAndTestError(trainingSetFile, ValidationSetFile, ValidationSetResultFile, testSetFile, testSetAnswerFile, alpha, epsilon, gamma, quadratic, ridge, rounded): x = gam.getAdjMat(trainingSetFile, quadratic) ## getAdjMatrix1.getmatrix returns a tuple std = x[0] mean = x[1] matrix = x[2] thetas = stochasticRegression(matrix, alpha, gamma, epsilon, ridge) print "...\nPredicting Validation Set..." predictedValues = predictValue(ValidationSetFile, thetas, quadratic, std, mean, "predictedValResult.csv", rounded) error = reportError(predictedValues, ValidationSetResultFile) print "The average squared error for the Validation Set is ", error print "...\nPredicting test Set..." predictedValuesT = predictValue(testSetFile, thetas, quadratic, std, mean, "predictedTestResult.csv", rounded) errorT = reportError(predictedValuesT, testSetAnswerFile) print "The average squared error status of this model is ", errorT thetas = np.insert(thetas, thetas.shape[0], alpha) thetas = np.insert(thetas, thetas.shape[0], gamma) thetas = np.insert(thetas, thetas.shape[0], error) thetas = np.insert(thetas, thetas.shape[0], errorT) return thetas
