def __init__(self, lexicon, C=1, num_features=100):
self.training_set = None
self.classes = None
self.test_set = None
self.results = None
self.kernel = ker.Linear()
self.C = C
self.feature_data = PATH + "/learning/stored/feature.data"
self.label_data = PATH + "/learning/stored/svm_label.data"
self.lexicon = lexicon
self.num_features = len(self.lexicon.words.keys())
try:
print "Loading existing SVM..."
features = pickle.load(open(self.feature_data))
labels = pickle.load(open(self.label_data))
sparsedata = SparseDataSet(features, L=labels)
self.svm_classifier = loadSVM(PATH + "/learning/stored/svm.classifier",sparsedata)
except Exception as e:
print e
print "Existing SVM not found!"
self.svm_classifier = svm.SVM(self.kernel)
self.accuracy = None
self.predicted_labels = None
score = featsel.FeatureScore('golub')
self.filter = featsel.Filter(score)
self.feature_selector = FeatureSelect(self.svm_classifier, self.filter)
self.chain = Chain([self.feature_selector, self.svm_classifier])
def __init__(self, lexicon, C=1, num_features=100):
self.training_set = None
self.classes = None
self.test_set = None
self.results = None
self.kernel = ker.Linear()
self.C = C
self.feature_data = PATH + "/learning/stored/feature.data"
self.label_data = PATH + "/learning/stored/svm_label.data"
self.lexicon = lexicon
self.num_features = len(self.lexicon.words.keys())
try:
print "Loading existing SVM..."
features = pickle.load(open(self.feature_data))
labels = pickle.load(open(self.label_data))
sparsedata = SparseDataSet(features, L=labels)
self.svm_classifier = loadSVM(PATH + "/learning/stored/svm.classifier", sparsedata)
except Exception as e:
print e
print "Existing SVM not found!"
self.svm_classifier = svm.SVM(self.kernel)
self.accuracy = None
self.predicted_labels = None
score = featsel.FeatureScore("golub")
self.filter = featsel.Filter(score)
self.feature_selector = FeatureSelect(self.svm_classifier, self.filter)
self.chain = Chain([self.feature_selector, self.svm_classifier])
def cross_validate(X,y):
# Ask what percentage of the data should be trained
p = 0
while p < 50 or p > 100:
p = raw_input("Enter percentage of data to train (between 50 and 75): ")
p = float(p);
K = int(floor(float(p)*len(X)/100))
X1 = X[0:K]
y1 = y[0:K]
#Load data into PyML's vector objects, then train set
data = VectorDataSet(X1,L=y1)
s = SVM()
s.train(data)
s.save("cross_validating")
#Now check the other data
X2 = X[K+1:-1]
y2 = y[K+1:-1]
#Load our training data
loadedSVM = loadSVM("cross_validating",data)
testData = VectorDataSet(X2,L=y2)
r = loadedSVM.test(testData)
print r
#Delete the data now that we're done with it
os.system("rm cross_validating")
def on_load_clicked(self, widget):
filen = self.getFilenameToRead("Load Classifier",filter='svm')
if filen is not None:
#db = shelve.open(filen)
#if db.has_key('clssfr'):
# self.clssfr = db['clssfr']
#else:
# self.showMessage("Cannot find a classifier!")
#db.close()
#with open(filen, 'wb') as f:
# self.clssfr = pickle.load(f)
datfn = self.getFilenameToRead("Open Training Data",filter='mat')
if datfn is not None:
data = ml.VectorDataSet(datfn,labelsColumn=0)
self.clssfr = loadSVM(filen,data) ## Why do I need to feed data ???
#self.clssfr = loadSVM(filen,None) ## edited PyML for this
# classifier has been loaded. need to update button status
self.setDisabledBtns()
self.showMessage("The classifier has been loaded!")
def on_load_clicked(self, widget):
filen = self.getFilenameToRead("Load Classifier", filter='svm')
if filen is not None:
#db = shelve.open(filen)
#if db.has_key('clssfr'):
# self.clssfr = db['clssfr']
#else:
# self.showMessage("Cannot find a classifier!")
#db.close()
#with open(filen, 'wb') as f:
# self.clssfr = pickle.load(f)
datfn = self.getFilenameToRead("Open Training Data", filter='mat')
if datfn is not None:
data = ml.VectorDataSet(datfn, labelsColumn=0)
self.clssfr = loadSVM(filen,
data) ## Why do I need to feed data ???
#self.clssfr = loadSVM(filen,None) ## edited PyML for this
# classifier has been loaded. need to update button status
self.setDisabledBtns()
self.showMessage("The classifier has been loaded!")
def test(component="svm", **args):
container = "SparseDataSet"
if "container" in args:
container = args["container"]
try:
DataSet = getattr(vectorDatasets, container)
except:
raise ValueError, "wrong container " + container
results = {}
comp = "general"
if component == "all" or component == comp:
s = svm.SVM()
results = {}
d = DataSet(heartdatafile, labelsColumn=0)
s.train(d)
s.test(d)
s = svm.SVM()
s.stratifiedCV(d)
print "starting aggregate****************"
d2 = Aggregate([d, d])
print "end aggregate"
r = s.stratifiedCV(d2)
d.attachKernel("polynomial")
s.cv(d)
d.attachKernel("linear")
s = svm.SVM()
s.train(d)
s.train(d, saveSpace=False)
s.save("tmp")
loaded = svm.loadSVM("tmp", datasetClass=DataSet)
r = loaded.test(d)
d.attachKernel("gaussian", gamma=0.01)
s.train(d, saveSpace=False)
s.save("tmp")
loaded = svm.loadSVM("tmp", datasetClass=DataSet, labelsColumn=1)
r = loaded.test(d)
os.remove("tmp")
d = DataSet(numpy.random.randn(100, 10))
d = DataSet([[1, 2], [2, 3]])
d = SequenceData(["asa", "ben", "hur"])
comp = "svm"
if component == "all" or component == comp:
d = DataSet(heartdatafile, labelsColumn=0)
results[comp] = []
d.attachKernel("polynomial")
s = svm.SVM()
results[comp].append(s.cv(d, saveSpace=True))
d.attachKernel("linear")
results[comp].append(s.cv(d))
comp = "kernelData"
if component == "all" or component == comp:
d = DataSet(heartdatafile, labelsColumn=0)
results[comp] = []
kdata = KernelData("heart.kernel", gistFormat=True)
kdata.attachLabels(d.labels)
s = svm.SVM()
results[comp].append(s.cv(kdata))
kdata.attachKernel("gaussian", gamma=0.1)
results[comp].append(s.cv(kdata))
comp = "normalization"
if component == "all" or component == comp:
results[comp] = []
data = DataSet(heartdatafile, labelsColumn=0)
data.attachKernel("polynomial", degree=4, normalization="dices")
s = svm.SVM()
results[comp].append(s.cv(data))
comp = "svr"
if component == "all" or component == comp:
d = DataSet(heartdatafile, labelsColumn=0, numericLabels=True)
results[comp] = []
s = svm.SVR()
# results[comp].append(
# s.cv(d, saveSpace = True))
# results[comp].append(
# s.trainTest(d, range(150), range(151, 250)))
results[comp].append(s.cv(d))
comp = "save"
if component == "all" or component == comp:
results[comp] = []
s = svm.SVM()
data = DataSet(heartdatafile, labelsColumn=0)
import tempfile
tmpfile = tempfile.mktemp()
r = s.cv(data)
r.save(tmpfile)
r = assess.loadResults(tmpfile)
results["save"].append(r)
r = s.nCV(data)
r.save(tmpfile)
results["save"].append(assess.loadResults(tmpfile))
r = {}
for i in range(10):
r[i] = s.cv(data)
assess.saveResultObjects(r, tmpfile)
r = assess.loadResults(tmpfile)
comp = "classifiers"
if component == "all" or component == comp:
d = DataSet(heartdatafile, labelsColumn=0)
results[comp] = []
cl = knn.KNN()
results[comp].append(cl.stratifiedCV(d))
print "testing ridge regression"
ridge = ridgeRegression.RidgeRegression()
results[comp].append(ridge.cv(d))
comp = "platt"
if component == "all" or component == "platt":
results[comp] = []
d = DataSet(heartdatafile, labelsColumn=0)
p = platt.Platt2(svm.SVM())
results[comp].append(p.stratifiedCV(d))
comp = "multi"
if component == "all" or component == comp:
results[comp] = []
d = DataSet(irisdatafile, labelsColumn=-1)
mc = multi.OneAgainstOne(svm.SVM())
results[comp].append(mc.cv(d))
d = DataSet(irisdatafile, labelsColumn=-1)
mc = multi.OneAgainstRest(svm.SVM())
results[comp].append(mc.cv(d))
mc = multi.OneAgainstRest(svm.SVM())
d.attachKernel("poly")
results[comp].append(mc.cv(d))
d.attachKernel("linear")
mc = multi.OneAgainstRest(svm.SVM())
# kdata = datafunc.KernelData('iris.linear.kernel',
# labelsFile = 'irisY.csv', labelsColumn = 0, gistFormat = True)
# results[comp].append(mc.cv(kdata))
comp = "featsel"
if component == "all" or component == comp:
results[comp] = []
s = svm.SVM()
d = DataSet(yeastdatafile, labelsColumn=0)
d2 = labels.oneAgainstRest(d, "2")
results[comp].append(s.stratifiedCV(d2))
# feature selection using RFE
m = composite.FeatureSelect(s, featsel.RFE())
results[comp].append(m.stratifiedCV(d2, 3))
fs = featsel.FeatureScore("golub")
f = featsel.Filter(fs, sigma=2)
f = featsel.Filter(fs, numFeatures=20)
m = composite.FeatureSelect(s, f)
results[comp].append(m.stratifiedCV(d2, 3))
# same thing but with a Chain:
c = composite.Chain([f, s])
# r = c.stratifiedCV (d2)
comp = "modelSelection"
if component == "all" or component == comp:
results[comp] = []
s = svm.SVM()
d = DataSet(heartdatafile, labelsColumn=0)
p = modelSelection.ParamGrid(svm.SVM(ker.Polynomial()), "C", [0.1, 1, 10, 100], "kernel.degree", [2, 3, 4])
p = modelSelection.ParamGrid(svm.SVM(ker.Gaussian()), "C", [0.1, 1, 10, 100], "kernel.gamma", [0.01, 0.1, 1])
# p = modelSelection.Param(svm.SVM(), 'C', [0.1, 1, 10, 100])
m = modelSelection.ModelSelector(p, measure="roc", foldsToPerform=2)
m = modelSelection.ModelSelector(p)
# m = modelSelection.SVMselect()
results[comp].append(m.cv(d))
return results
def test(component='svm', **args):
container = 'SparseDataSet'
if 'container' in args:
container = args['container']
try:
DataSet = getattr(vectorDatasets, container)
except:
raise ValueError, 'wrong container ' + container
results = {}
comp = 'general'
if component == 'all' or component == comp:
s = svm.SVM()
results = {}
d = DataSet(heartdatafile, labelsColumn=0)
s.train(d)
s.test(d)
s = svm.SVM()
s.stratifiedCV(d)
print 'starting aggregate****************'
d2 = Aggregate([d, d])
print 'end aggregate'
r = s.stratifiedCV(d2)
d.attachKernel('polynomial')
s.cv(d)
d.attachKernel('linear')
s = svm.SVM()
s.train(d)
s.train(d, saveSpace=False)
s.save("tmp")
loaded = svm.loadSVM("tmp", datasetClass=DataSet)
r = loaded.test(d)
d.attachKernel('gaussian', gamma=0.01)
s.train(d, saveSpace=False)
s.save("tmp")
loaded = svm.loadSVM("tmp", datasetClass=DataSet, labelsColumn=1)
r = loaded.test(d)
os.remove('tmp')
d = DataSet(numpy.random.randn(100, 10))
d = DataSet([[1, 2], [2, 3]])
d = SequenceData(['asa', 'ben', 'hur'])
comp = 'svm'
if component == 'all' or component == comp:
d = DataSet(heartdatafile, labelsColumn=0)
results[comp] = []
d.attachKernel('polynomial')
s = svm.SVM()
results[comp].append(s.cv(d, saveSpace=True))
d.attachKernel('linear')
results[comp].append(s.cv(d))
comp = 'kernelData'
if component == 'all' or component == comp:
d = DataSet(heartdatafile, labelsColumn=0)
results[comp] = []
kdata = KernelData('heart.kernel', gistFormat=True)
kdata.attachLabels(d.labels)
s = svm.SVM()
results[comp].append(s.cv(kdata))
kdata.attachKernel('gaussian', gamma=0.1)
results[comp].append(s.cv(kdata))
comp = 'normalization'
if component == 'all' or component == comp:
results[comp] = []
data = DataSet(heartdatafile, labelsColumn=0)
data.attachKernel('polynomial', degree=4, normalization='dices')
s = svm.SVM()
results[comp].append(s.cv(data))
comp = 'svr'
if component == 'all' or component == comp:
d = DataSet(heartdatafile, labelsColumn=0, numericLabels=True)
results[comp] = []
s = svm.SVR()
#results[comp].append(
# s.cv(d, saveSpace = True))
#results[comp].append(
# s.trainTest(d, range(150), range(151, 250)))
results[comp].append(s.cv(d))
comp = 'save'
if component == 'all' or component == comp:
results[comp] = []
s = svm.SVM()
data = DataSet(heartdatafile, labelsColumn=0)
import tempfile
tmpfile = tempfile.mktemp()
r = s.cv(data)
r.save(tmpfile)
r = assess.loadResults(tmpfile)
results['save'].append(r)
r = s.nCV(data)
r.save(tmpfile)
results['save'].append(assess.loadResults(tmpfile))
r = {}
for i in range(10):
r[i] = s.cv(data)
assess.saveResultObjects(r, tmpfile)
r = assess.loadResults(tmpfile)
comp = 'classifiers'
if component == 'all' or component == comp:
d = DataSet(heartdatafile, labelsColumn=0)
results[comp] = []
cl = knn.KNN()
results[comp].append(cl.stratifiedCV(d))
print 'testing ridge regression'
ridge = ridgeRegression.RidgeRegression()
results[comp].append(ridge.cv(d))
comp = 'platt'
if component == 'all' or component == 'platt':
results[comp] = []
d = DataSet(heartdatafile, labelsColumn=0)
p = platt.Platt2(svm.SVM())
results[comp].append(p.stratifiedCV(d))
comp = 'multi'
if component == 'all' or component == comp:
results[comp] = []
d = DataSet(irisdatafile, labelsColumn=-1)
mc = multi.OneAgainstOne(svm.SVM())
results[comp].append(mc.cv(d))
d = DataSet(irisdatafile, labelsColumn=-1)
mc = multi.OneAgainstRest(svm.SVM())
results[comp].append(mc.cv(d))
mc = multi.OneAgainstRest(svm.SVM())
d.attachKernel('poly')
results[comp].append(mc.cv(d))
d.attachKernel('linear')
mc = multi.OneAgainstRest(svm.SVM())
#kdata = datafunc.KernelData('iris.linear.kernel',
# labelsFile = 'irisY.csv', labelsColumn = 0, gistFormat = True)
#results[comp].append(mc.cv(kdata))
comp = 'featsel'
if component == 'all' or component == comp:
results[comp] = []
s = svm.SVM()
d = DataSet(yeastdatafile, labelsColumn=0)
d2 = labels.oneAgainstRest(d, '2')
results[comp].append(s.stratifiedCV(d2))
# feature selection using RFE
m = composite.FeatureSelect(s, featsel.RFE())
results[comp].append(m.stratifiedCV(d2, 3))
fs = featsel.FeatureScore('golub')
f = featsel.Filter(fs, sigma=2)
f = featsel.Filter(fs, numFeatures=20)
m = composite.FeatureSelect(s, f)
results[comp].append(m.stratifiedCV(d2, 3))
# same thing but with a Chain:
c = composite.Chain([f, s])
#r = c.stratifiedCV (d2)
comp = 'modelSelection'
if component == 'all' or component == comp:
results[comp] = []
s = svm.SVM()
d = DataSet(heartdatafile, labelsColumn=0)
p = modelSelection.ParamGrid(svm.SVM(ker.Polynomial()), 'C',
[0.1, 1, 10, 100], 'kernel.degree',
[2, 3, 4])
p = modelSelection.ParamGrid(svm.SVM(ker.Gaussian()), 'C',
[0.1, 1, 10, 100], 'kernel.gamma',
[0.01, 0.1, 1])
#p = modelSelection.Param(svm.SVM(), 'C', [0.1, 1, 10, 100])
m = modelSelection.ModelSelector(p, measure='roc', foldsToPerform=2)
m = modelSelection.ModelSelector(p)
#m = modelSelection.SVMselect()
results[comp].append(m.cv(d))
return results
if line.rstrip() == 'True':
temp = 'True'
else:
temp = 'False'
y.append(temp)
## Now it's time to load our data into PyML's vector objects
data = VectorDataSet(X2,L=y2)
#Create SVM object, then train our set
s = SVM()
s.train(data)
s.save("freePizza")
## Yay!
# Now to cross-validate the data; we first take the other set
X3 = X[fifth+1:-1];
y3 = y[fifth+1:-1];
print y3
#Load our training data
from PyML.classifiers.svm import loadSVM
loadedSVM = loadSVM("freePizza",data)
testData = VectorDataSet(X3,L=y3)
r = loadedSVM.test(testData)
print r
def svm_prediction(peptides, job_id, input_train="SVM_POS_NEG.fasta"):
"""
Makes a final prediction based on SVM training files.
This code is used for prediciton of blind datasets, based on the training
datasets of positives and negatives.
:param peptides: input peptides
:param job_id: random job id assigned prior to start predicting
:param input_train: input positive and negative examples used in training
:return: returns SVM scores for each inputed peptide
"""
print("Begin SVM")
# from methods import load_sqlite, store_sqlite
global PATH
global TMP_PATH
# suppress SVM output
devnull = open(os.devnull, 'w')
sys.stdout, sys.stderr = devnull, devnull
svm_scores = []
# query the database
# for peptide in peptides:
# try:
# score = load_sqlite(peptide, method="SVM", unique=True)
# svm_scores.append(score)
# except:
# pass
if len(peptides) == len(svm_scores):
pass
else:
# generate a svm input from the peptides
rand = job_id
input_svm = "%s_svm.fasta" % rand
output_tmp = open(os.path.join(TMP_PATH, input_svm), "w")
count = 0
for peptide in peptides:
count += 1
output_tmp.write("> %i label=%s\n%s\n" % (count, 1, peptide))
for peptide in peptides:
count += 1
output_tmp.write("> %i label=%s\n%s\n" % (count, -1, peptide))
output_tmp.close()
# outputs
model_svm = "%s_svm_model.txt" % rand
# train data
train_data = SequenceData(os.path.join(PATH, input_train),
mink=1,
maxk=1,
maxShift=0,
headerHandler=svm_process_header)
train_data.attachKernel('cosine')
cval = 1
s = SVM(C=cval)
s.train(train_data)
s.save(os.path.join(TMP_PATH, model_svm))
# load trained SVM
loaded_svm = loadSVM(os.path.join(TMP_PATH, model_svm), train_data)
# test data
test_data = SequenceData(os.path.join(TMP_PATH, input_svm),
mink=1,
maxk=1,
maxShift=0,
headerHandler=svm_process_header)
test_data.attachKernel('cosine')
results = loaded_svm.test(test_data)
# print results out
output_svm = "%s_svm.txt" % rand
results.toFile(os.path.join(TMP_PATH, output_svm))
# load results process output (positives + negatives)
infile = open(os.path.join(TMP_PATH, output_svm), "r")
inlines = infile.readlines()
infile.close()
scores = list()
for line in inlines:
line = line.rstrip("\r\n")
try:
entry = int(line.split("\t")[0])
score = float(line.split("\t")[1])
label = int(line.split("\t")[3])
if label != "-1":
scores.append([entry, score])
except:
pass
# order list
sorted_scores = sorted(scores, key=lambda scores: scores[0])
svm_scores = list()
for score in sorted_scores:
svm_score = score[1]
svm_scores.append(svm_score)
# remove the temporary model files and results
try:
os.remove(os.path.join(TMP_PATH, input_svm))
os.remove(os.path.join(TMP_PATH, model_svm))
os.remove(os.path.join(TMP_PATH, output_svm))
except:
pass
# save the peptides in db
# for peptide, score in zip(peptides, svm_scores):
# store_sqlite(peptide, method="SVM", information=score, save=True)
# restore normal output
sys.stdout = sys.__stdout__
sys.stderr = sys.__stderr__
print("End SVM")
return svm_scores
