def genMusic(genScore, args): samples = featureManager.loadJson(config.getGenOutFeatFilename(args)) sample = samples[0] #only one score, to match the format of trainFeat perfFeats = sample['perfFeats'] outScore = featureManager.applyFeats(genScore, perfFeats) musicOutputFunc = getMusicOutputFunc() musicOutputFunc(outScore, args)
def gen(self, args):
genFeats = featureManager.loadJson(config.getGenInFeatFilename(args))
genFeat = genFeats[0]
scoreName = genFeat['name']
scoreFeats = genFeat['scoreFeats']
#wrap genFeats in [] to match data structure in train
lines = self.formatLineDirect([genFeat])
#logging.printDebug(lines)
allLines = map(lambda l: "0 " + l, lines)
#logging.printDebug(allLines)
#TODO: collect filenames
SVMGenInputFilename = self.getGenInputFilename(args)
with open(SVMGenInputFilename, 'w') as f:
f.writelines(map(lambda x:x+"\n", allLines))
for featName in config.perfFeatsList:
cmd = [self.genBinPath]
cmd.append(SVMGenInputFilename)
singleModelFilename= self.getSingleModelFilename(args, featName)
cmd.append(singleModelFilename)
perfFeatFilename= self.getGenOutputFilename(args, featName)
cmd.append(perfFeatFilename)
#logging.printDebug(" ".join(cmd))
subprocess.call(" ".join(cmd), shell=True)
#TODO: read all perfFeatFilename, transform to dict and return
perfFeats = {}
for featName in config.perfFeatsList:
#TODO: put all filename definition in one place
#perfFeatFilename= self.getGenOutputFilename(args, featName)
with open(perfFeatFilename, 'r') as f:
lines = f.readlines()
#dequantize
q = quantizer.getQuantizerObj(self.getQuantizeFilename(args, featName))
realVals= q.dequantize(lines)
#feat = map(float, lines)
perfFeats[featName]=realVals
#logging.printDebug(perfFeats)
return featureManager.formatFeatFile(scoreName, {}, perfFeats)
def gen(self, args):
genFeats = featureManager.loadJson(config.getGenInFeatFilename(args))
genFeat = genFeats[0]
scoreName = genFeat['name']
scoreFeats = genFeat['scoreFeats']
#wrap genFeats in [] to match data structure in train
lines = self.formatLineDirect([genFeat])
#logging.printDebug(lines)
allLines = map(lambda l: "0 " + l, lines)
#logging.printDebug(allLines)
#TODO: collect filenames
SVMGenInputFilename = self.getGenInputFilename(args)
with open(SVMGenInputFilename, 'w') as f:
f.writelines(map(lambda x: x + "\n", allLines))
for featName in config.perfFeatsList:
cmd = [self.genBinPath]
cmd.append(SVMGenInputFilename)
singleModelFilename = self.getSingleModelFilename(args, featName)
cmd.append(singleModelFilename)
perfFeatFilename = self.getGenOutputFilename(args, featName)
cmd.append(perfFeatFilename)
#logging.printDebug(" ".join(cmd))
subprocess.call(" ".join(cmd), shell=True)
#TODO: read all perfFeatFilename, transform to dict and return
perfFeats = {}
for featName in config.perfFeatsList:
#TODO: put all filename definition in one place
#perfFeatFilename= self.getGenOutputFilename(args, featName)
with open(perfFeatFilename, 'r') as f:
lines = f.readlines()
#dequantize
q = quantizer.getQuantizerObj(
self.getQuantizeFilename(args, featName))
realVals = q.dequantize(lines)
#feat = map(float, lines)
perfFeats[featName] = realVals
#logging.printDebug(perfFeats)
return featureManager.formatFeatFile(scoreName, {}, perfFeats)
def train(self, args):
#ex filenames:
# trainFeat.json -> [quantizer] -> <perfFeatName>.train.dat
# | |
# +-> <perfFeatName>.quant |
# +-------------------------------------------------------+
# +-> [svm^hmm] -> <perfFeatName>.model.bin
trainFeats = featureManager.loadJson(
config.getTrainInFeatFilename(args))
perfFeats = collectPerfFeats(trainFeats)
#logging.printDebug(perfFeats)
#for each perfFeature, a svmFeatFilename (quantized feat for svm^hmm)
#and singleModelFilename (model file for svm^hmm) will be saved
#scoreFeatLines = self.formatLine(scoreFeats)
scoreFeatLines = self.formatLineDirect(trainFeats)
for pkey, pval in perfFeats.items():
allLines = [("# " + pkey)]
#TODO:quantize pval
q = quantizer.getQuantizerObj(self.getQuantizeFilename(args, pkey))
quantizedVal = q.quantize(pval)
lines = zip(map(str, quantizedVal), scoreFeatLines)
#logging.printDebug(lines)
allLines = map(lambda l: " ".join(l), lines)
#logging.printDebug(allLines)
svmFeatFilename = self.getTrainInputFilename(args, pkey)
with open(svmFeatFilename, 'w') as f:
f.writelines(map(lambda x: x + "\n", allLines))
cmd = [self.trainBinPath]
if (config.svmhmm_c != None):
cmd.append("-c " + str(config.svmhmm_c))
else:
cmd.append("-c 0.01")
cmd.append(svmFeatFilename)
singleModelFilename = self.getSingleModelFilename(args, pkey)
cmd.append(singleModelFilename)
#logging.printDebug(" ".join(cmd))
subprocess.call(" ".join(cmd), shell=True)
def train(self, args):
#ex filenames:
# trainFeat.json -> [quantizer] -> <perfFeatName>.train.dat
# | |
# +-> <perfFeatName>.quant |
# +-------------------------------------------------------+
# +-> [svm^hmm] -> <perfFeatName>.model.bin
trainFeats = featureManager.loadJson(config.getTrainInFeatFilename(args))
perfFeats = collectPerfFeats(trainFeats)
#logging.printDebug(perfFeats)
#for each perfFeature, a svmFeatFilename (quantized feat for svm^hmm)
#and singleModelFilename (model file for svm^hmm) will be saved
#scoreFeatLines = self.formatLine(scoreFeats)
scoreFeatLines = self.formatLineDirect(trainFeats)
for pkey, pval in perfFeats.items():
allLines = [("# " + pkey)]
#TODO:quantize pval
q = quantizer.getQuantizerObj(self.getQuantizeFilename(args, pkey))
quantizedVal = q.quantize(pval)
lines = zip(map(str, quantizedVal), scoreFeatLines)
#logging.printDebug(lines)
allLines = map(lambda l:" ".join(l), lines)
#logging.printDebug(allLines)
svmFeatFilename = self.getTrainInputFilename(args, pkey)
with open(svmFeatFilename, 'w') as f:
f.writelines(map(lambda x:x+"\n", allLines))
cmd = [self.trainBinPath]
if (config.svmhmm_c != None):
cmd.append("-c " + str(config.svmhmm_c))
else:
cmd.append("-c 0.01")
cmd.append(svmFeatFilename)
singleModelFilename = self.getSingleModelFilename(args,pkey)
cmd.append(singleModelFilename)
#logging.printDebug(" ".join(cmd))
subprocess.call(" ".join(cmd), shell=True)
def train(self, trainFeatFilename, modelFilename):
trainFeats = featureManager.loadJson(trainFeatFilename)
scoreFeats = collectScoreFeats(trainFeats)
perfFeats = collectPerfFeats(trainFeats)
#data = r('data.frame(scoreFeats, perfFeats)')
formula = 'cbind(' + ','.join(config.perfFeatsList) +')'
formula += '~'
formula += '*'.join(config.scoreFeatsList)
#logging.printDebug(formula)
r.assign('fstr', formula)
r.assign('scoreFeats',scoreFeats)
r.assign('perfFeats',perfFeats)
r('f <- formula(fstr)')
r('scoreFeats <- data.frame(scoreFeats)')
r('perfFeats <- data.frame(perfFeats)')
if config.DEBUG:
r('str(perfFeats)')
r('str(scoreFeats)')
r('str(f)')
r.assign('modelFilename',modelFilename)
r.source('modelMultiLinearRegress.R')
r('trainMultiLinearRegress(f, perfFeats, scoreFeats, modelFilename)')
def gen(self, genFeatFilename, modelFilename):
genFeats = featureManager.loadJson(genFeatFilename)
scoreFeats = genFeats['scoreFeats']
#logging.printDebug(scoreFeats)
r.assign('scoreFeats',scoreFeats)
r.assign('modelFilename',modelFilename)
r('scoreFeats <- data.frame(scoreFeats)')
if config.DEBUG:
r('str(scoreFeats)')
r('save(scoreFeats, file="../output/debug_genFeat.rda")')
r.source('modelMultiLinearRegress.R')
perfFeats = r('perfFeats<-genMultiLinearRegress(scoreFeats, modelFilename)')
if config.DEBUG: r('str(perfFeats)')
#workaround: if only one perfFeat is used. predict will return a bad format data.frame without proper perfFeatName
#logging.printDebug('PerfFeatures from genModel')
#logging.printDebug(perfFeats)
#logging.printDebug(perfFeats.items())
if len(perfFeats) == 1:
newPerfFeats = {}
for key, val in perfFeats.items():
newPerfFeats[config.perfFeatsList[0]] = val
return newPerfFeats
return perfFeats
def train(self, trainFeatFilename, modelFilename):
trainFeats = featureManager.loadJson(trainFeatFilename)
scoreFeats = collectScoreFeats(trainFeats)
perfFeats = collectPerfFeats(trainFeats)
#data = r('data.frame(scoreFeats, perfFeats)')
formula = 'cbind(' + ','.join(config.perfFeatsList) + ')'
formula += '~'
formula += '*'.join(config.scoreFeatsList)
#logging.printDebug(formula)
r.assign('fstr', formula)
r.assign('scoreFeats', scoreFeats)
r.assign('perfFeats', perfFeats)
r('f <- formula(fstr)')
r('scoreFeats <- data.frame(scoreFeats)')
r('perfFeats <- data.frame(perfFeats)')
if config.DEBUG:
r('str(perfFeats)')
r('str(scoreFeats)')
r('str(f)')
r.assign('modelFilename', modelFilename)
r.source('modelMultiLinearRegress.R')
r('trainMultiLinearRegress(f, perfFeats, scoreFeats, modelFilename)')
def gen(self, genFeatFilename, modelFilename):
genFeats = featureManager.loadJson(genFeatFilename)
scoreFeats = genFeats['scoreFeats']
#logging.printDebug(scoreFeats)
r.assign('scoreFeats', scoreFeats)
r.assign('modelFilename', modelFilename)
r('scoreFeats <- data.frame(scoreFeats)')
if config.DEBUG:
r('str(scoreFeats)')
r('save(scoreFeats, file="../output/debug_genFeat.rda")')
r.source('modelMultiLinearRegress.R')
perfFeats = r(
'perfFeats<-genMultiLinearRegress(scoreFeats, modelFilename)')
if config.DEBUG: r('str(perfFeats)')
#workaround: if only one perfFeat is used. predict will return a bad format data.frame without proper perfFeatName
#logging.printDebug('PerfFeatures from genModel')
#logging.printDebug(perfFeats)
#logging.printDebug(perfFeats.items())
if len(perfFeats) == 1:
newPerfFeats = {}
for key, val in perfFeats.items():
newPerfFeats[config.perfFeatsList[0]] = val
return newPerfFeats
return perfFeats
#main
parser = argparse.ArgumentParser()
parser.add_argument(
"dir", #nargs="1" ,
help="Path to all feature files.",
default=defaultDir)
args = parser.parse_args()
filenames = os.listdir(args.dir)
realPerfFilnames = fnmatch.filter(filenames, realPerfExt)
genPerfFilnames = fnmatch.filter(filenames, genPerfExt)
lines = []
#prepare col names
feat = featureManager.loadJson(args.dir + realPerfFilnames[0])
lines += parseColNames(feat)
for realPerfFilname in realPerfFilnames:
if (DEBUG):
lines += ["# " + realPerfFilname]
realFeats = featureManager.loadJson(args.dir + realPerfFilname)
lines += parseLines(realFeats, "real")
for genPerfFilname in genPerfFilnames:
if (DEBUG):
lines += ["# " + genPerfFilname]
genFeats = featureManager.loadJson(args.dir + genPerfFilname)
lines += parseLines(genFeats, "gen")
for line in lines:
import model
import config
import featureManager
trainFeats = featureManager.loadJson(config.defaultTrainFeatsFilename)
print(model.collectScoreFeats(trainFeats))
print(model.collectPerfFeats(trainFeats))
model = model.trainMultiLinearRegress(trainFeats, config.defaultModelFilename)
print(model)
def dummyFunc():
print("this is a function for type(dummyFunc) comparison")
#for mode in ['train', 'gen']:
# print(model.getModelFunc(mode))
# assert type(model.getModelFunc(mode)) == type(dummyFunc), 'getModelFunc() returned a non-function object'
#
import model
import config
import featureManager
trainFeats = featureManager.loadJson(config.defaultTrainFeatsFilename)
print(model.collectScoreFeats(trainFeats))
print(model.collectPerfFeats(trainFeats))
model = model.trainMultiLinearRegress(trainFeats, config.defaultModelFilename)
print(model)
def dummyFunc():
print("this is a function for type(dummyFunc) comparison")
#for mode in ['train', 'gen']:
# print(model.getModelFunc(mode))
# assert type(model.getModelFunc(mode)) == type(dummyFunc), 'getModelFunc() returned a non-function object'
#
return lines
# main
parser = argparse.ArgumentParser()
parser.add_argument("dir", help="Path to all feature files.", default=defaultDir) # nargs="1" ,
args = parser.parse_args()
filenames = os.listdir(args.dir)
realPerfFilnames = fnmatch.filter(filenames, realPerfExt)
genPerfFilnames = fnmatch.filter(filenames, genPerfExt)
lines = []
# prepare col names
feat = featureManager.loadJson(args.dir + realPerfFilnames[0])
lines += parseColNames(feat)
for realPerfFilname in realPerfFilnames:
if DEBUG:
lines += ["# " + realPerfFilname]
realFeats = featureManager.loadJson(args.dir + realPerfFilname)
lines += parseLines(realFeats, "real")
for genPerfFilname in genPerfFilnames:
if DEBUG:
lines += ["# " + genPerfFilname]
genFeats = featureManager.loadJson(args.dir + genPerfFilname)
lines += parseLines(genFeats, "gen")