def main(): Renascence.setStreamPath("./") Renascence.setLibPath("./") producer = Renascence.init(["func.xml"]) print producer.listAllFunctions() print producer.listAllTypes() x0 = producer.load('TrBmp', "input.jpg") x1 = producer.load('TrBmp', "input_sharp.jpg") x2 = producer.load('TrBmp', "output.jpg") filterMatrix = producer.build("FR(x0, x1)").run(producer.merge(x2, x1)) print filterMatrix.dump() [Trainner, BestValue] = producer.train("S(ADF(Treator,x0,x1))", producer.merge(x0, x1), 5, cacheFile='temp.txt', postFormula='FIT(x0, x1)', postExtraInput=x2) formula = Trainner.ADF("") print 'Formula: ', formula print 'Parameters: ', Trainner.parameters() print BestValue predictor = producer.build(formula, Trainner.parameters()) y_p = predictor.run(producer.merge(x0, x1)) y_p.save("output/output_trained.jpg")
def main(): producer = Renascence.init(["./libAbstract_learning.xml"]) gd = producer.load('ALGradientMethod', './res/cnn/lenet.json') #gd = producer.load('ALGradientMethod', './res/cnn/softmax.json') trainMerge = producer.load('ALFloatMatrix', '../../machine_exam/handset/train.txt') trainX = producer.build( 'MatrixLinear(MatrixCrop(x0, x1, x2), x3, x4)').run( producer.merge(trainMerge, 1.0, -1.0, 1.0 / 255.0, 0.0)) trainX.save('output/handset.trainX.txt') trainY = producer.build('MatrixCrop(x0, x1, x2)').run( producer.merge(trainMerge, 0.0, 0.0)) trainY.save('output/handset.trainY.txt') p = producer.build('ParameterInit(x0)').run(gd) print trainX, trainY, gd, p mergeM = producer.build('GDMatrixPrepare(x0, x1, x2)').run( producer.merge(trainX, trainY, gd)) mergeM.save('output/handset.m') p = producer.build('GDCompute(x0, x1, x2)').run( producer.merge(mergeM, gd, p)) p.save('output/handset.p') #p = producer.load('ALFloatMatrix', 'output/pieces/parameters/ALFloatMatrix_0') #print p predictX = producer.load('ALFloatMatrix', '../../machine_exam/handset/test.txt') print predictX predictY = producer.build( 'Classify(GDPredictorLoad(x0, x1), MatrixLinear(x2,x3,x4))').run( producer.merge(gd, p, predictX, 1.0 / 255.0, 0.0)) predictY.save('output/handset.y')
def getPieceProducer(): producer = Renascence.init(["./libAbstract_learning.xml"]) print producer p_producer = Renascence.PieceFunctionProducer(producer, ['./parallel.xml'], ['Map-Reduce.xml']) print p_producer sub_p_producer = Renascence.PieceFunctionProducerParallel( p_producer, 'server') return sub_p_producer
def main(): Renascence.setStreamPath("./") Renascence.setLibPath("./") producer = Renascence.init(["func.xml"]) p_producer = Renascence.PieceFunctionProducer(producer, ['parallel.xml'], ['Map-Reduce.xml']) print p_producer.listType() sub_p_producer = Renascence.PieceFunctionProducerParallel(p_producer, 'server') print "sub_p_producer created" pfunction = sub_p_producer.createFunction('C(S(x0))', 'TrBmp') inputPieces = sub_p_producer.createInput('deps/Renascence/res/pictures/', 'TrBmp', [5]) tempPiece = pfunction.run([inputPieces]) outputPieces = sub_p_producer.createOutput('output/pythonTestParallel') sub_p_producer.copyPiece(tempPiece, outputPieces)
def main(): Renascence.setStreamPath("./") Renascence.setLibPath("./") producer = Renascence.init(["func.xml"]) p_producer = Renascence.PieceFunctionProducer(producer, ['mgpfunc.xml'], ['Map-Reduce.xml']) print p_producer.listType() sub_p_producer = Renascence.PieceFunctionProducerParallel(p_producer, 'thread') pfunction = sub_p_producer.createFunction('C(S(x0))', 'TrBmp') inputPieces = sub_p_producer.createInput('res/pictures/', 'TrBmp', [5]) outputPieces = sub_p_producer.createOutput('output/pythonTestParallel') tempPiece = pfunction.run([inputPieces]) sub_p_producer.copyPiece(tempPiece, outputPieces) input1 = inputPieces.read('0') input1.save('output/pythonTestParallel/input.jpg') tempPiece.write(input1, '0') print input1
def main(): Renascence.setStreamPath("../../") Renascence.setLibPath("../../") producer = Renascence.init(["func.xml"]) print producer.listAllFunctions() print producer.listAllTypes() x0 = producer.load('TrBmp', "input.jpg") x1 = producer.load('TrBmp', "input_sharp.jpg") x2 = producer.load('TrBmp', "output.jpg") x0x1x2 = producer.merge(x0, x1, x2) [Trainner, BestValue] = producer.train("FIT(ADF(Treator,x0,x1), x2)", x0x1x2, 500, cacheFile='temp.txt') formula = Trainner.ADF("Treator") print 'Formula: ', formula print 'Parameters: ', Trainner.parameters() print BestValue predictor = producer.build(formula) x0x1 = producer.merge(x0, x1) y_p = predictor.run(x0x1) u = producer.content(9.0) y_p.save("output/output_trained.jpg")
def main(): Renascence.setStreamPath("./") Renascence.setLibPath("./") producer = Renascence.init(["func.xml"]) print producer.listAllFunctions() print producer.listAllTypes() x0 = producer.load('TrBmp', "input.jpg") x1 = producer.load('TrBmp', "input_sharp.jpg") x2 = producer.load('TrBmp', "output.jpg") filterMatrix = producer.build("FR(x0, x1)").run(producer.merge(x2, x1)) print filterMatrix.dump() [Trainner, BestValue] = producer.train("S(ADF(Treator,x0,x1))", producer.merge(x0, x1), 5, cacheFile='temp.txt', postFormula='FIT(x0, x1)', postExtraInput=x2) formula = Trainner.ADF("") print 'Formula: ', formula print 'Parameters: ', Trainner.parameters() print BestValue predictor = producer.build(formula, Trainner.parameters()) y_p = predictor.run(producer.merge(x0,x1)) y_p.save("output/output_trained.jpg")
#!/usr/bin/python import Renascence producer = Renascence.init(["./libAbstract_learning.xml"]) print producer.listAllFunctions() print producer.listAllTypes() x0 = producer.load('ALFloatMatrix', './a1a.tra') #formula = 'CrossValidate(ADF(GodTrain), Labled(x0, x1))' #formula = 'C45Tree(x0)' formula = 'RandomForest(x0)' result = producer.build(formula).run(x0) result.save('./.tree')