def test_join_options(self):
"""
Tests the join_options method.
"""
self.assertEqual("", str(classes.join_options([])))
self.assertEqual(
"-t /some/where/test.arff",
str(classes.join_options(["-t", "/some/where/test.arff"])))
def main():
"""
Runs a datagenerator from the command-line. Calls JVM start/stop automatically.
Use -h to see all options.
"""
parser = argparse.ArgumentParser(
description='Executes a data generator from the command-line. Calls JVM start/stop automatically.')
parser.add_argument("-j", metavar="classpath", dest="classpath", help="additional classpath, jars/directories")
parser.add_argument("-X", metavar="heap", dest="heap", help="max heap size for jvm, e.g., 512m")
parser.add_argument("datagenerator", help="data generator classname, e.g., "
+ "weka.datagenerators.classifiers.classification.LED24")
parser.add_argument("option", nargs=argparse.REMAINDER, help="additional data generator options")
parsed = parser.parse_args()
jars = []
if parsed.classpath is not None:
jars = parsed.classpath.split(os.pathsep)
jvm.start(jars, max_heap_size=parsed.heap, packages=True)
logger.debug("Commandline: " + join_options(sys.argv[1:]))
try:
generator = DataGenerator(classname=parsed.datagenerator)
if len(parsed.option) > 0:
generator.options = parsed.option
DataGenerator.make_data(generator, parsed.option)
except Exception, e:
print(e)
def main():
"""
Runs a associator from the command-line. Calls JVM start/stop automatically.
Use -h to see all options.
"""
parser = argparse.ArgumentParser(
description='Executes an associator from the command-line. Calls JVM start/stop automatically.')
parser.add_argument("-j", metavar="classpath", dest="classpath", help="additional classpath, jars/directories")
parser.add_argument("-X", metavar="heap", dest="heap", help="max heap size for jvm, e.g., 512m")
parser.add_argument("-t", metavar="train", dest="train", required=True, help="training set file")
parser.add_argument("associator", help="associator classname, e.g., weka.associations.Apriori")
parser.add_argument("option", nargs=argparse.REMAINDER, help="additional associator options")
parsed = parser.parse_args()
jars = []
if parsed.classpath is not None:
jars = parsed.classpath.split(os.pathsep)
jvm.start(jars, max_heap_size=parsed.heap, packages=True)
logger.debug("Commandline: " + join_options(sys.argv[1:]))
try:
associator = Associator(classname=parsed.associator)
if len(parsed.option) > 0:
associator.options = parsed.option
loader = converters.loader_for_file(parsed.train)
data = loader.load_file(parsed.train)
associator.build_associations(data)
print(str(associator))
except Exception, e:
print(e)
def main():
"""
Runs a associator from the command-line. Calls JVM start/stop automatically.
Use -h to see all options.
"""
parser = argparse.ArgumentParser(
description='Executes an associator from the command-line. Calls JVM start/stop automatically.')
parser.add_argument("-j", metavar="classpath", dest="classpath", help="additional classpath, jars/directories")
parser.add_argument("-X", metavar="heap", dest="heap", help="max heap size for jvm, e.g., 512m")
parser.add_argument("-t", metavar="train", dest="train", required=True, help="training set file")
parser.add_argument("associator", help="associator classname, e.g., weka.associations.Apriori")
parser.add_argument("option", nargs=argparse.REMAINDER, help="additional associator options")
parsed = parser.parse_args()
jars = []
if parsed.classpath is not None:
jars = parsed.classpath.split(os.pathsep)
jvm.start(jars, max_heap_size=parsed.heap, packages=True)
logger.debug("Commandline: " + join_options(sys.argv[1:]))
try:
associator = Associator(classname=parsed.associator)
if len(parsed.option) > 0:
associator.options = parsed.option
loader = converters.loader_for_file(parsed.train)
data = loader.load_file(parsed.train)
associator.build_associations(data)
print(str(associator))
except Exception as e:
print(e)
finally:
jvm.stop()
def main():
"""
Runs a clusterer from the command-line. Calls JVM start/stop automatically.
Use -h to see all options.
"""
parser = argparse.ArgumentParser(
description='Performs clustering from the command-line. Calls JVM start/stop automatically.')
parser.add_argument("-j", metavar="classpath", dest="classpath", help="additional classpath, jars/directories")
parser.add_argument("-X", metavar="heap", dest="heap", help="max heap size for jvm, e.g., 512m")
parser.add_argument("-t", metavar="train", dest="train", required=True, help="training set file")
parser.add_argument("-T", metavar="test", dest="test", help="test set file")
parser.add_argument("-d", metavar="outmodel", dest="outmodel", help="model output file name")
parser.add_argument("-l", metavar="inmodel", dest="inmodel", help="model input file name")
parser.add_argument("-p", metavar="attributes", dest="attributes", help="attribute range")
parser.add_argument("-x", metavar="num folds", dest="numfolds", help="number of folds")
parser.add_argument("-s", metavar="seed", dest="seed", help="seed value for randomization")
parser.add_argument("-c", metavar="class index", dest="classindex", help="1-based class attribute index")
parser.add_argument("-g", metavar="graph", dest="graph", help="graph output file (if supported)")
parser.add_argument("clusterer", help="clusterer classname, e.g., weka.clusterers.SimpleKMeans")
parser.add_argument("option", nargs=argparse.REMAINDER, help="additional clusterer options")
parsed = parser.parse_args()
jars = []
if parsed.classpath is not None:
jars = parsed.classpath.split(os.pathsep)
params = []
if parsed.train is not None:
params.extend(["-t", parsed.train])
if parsed.test is not None:
params.extend(["-T", parsed.test])
if parsed.outmodel is not None:
params.extend(["-d", parsed.outmodel])
if parsed.inmodel is not None:
params.extend(["-l", parsed.inmodel])
if parsed.attributes is not None:
params.extend(["-p", parsed.attributes])
if parsed.numfolds is not None:
params.extend(["-x", parsed.numfolds])
if parsed.seed is not None:
params.extend(["-s", parsed.seed])
if parsed.classindex is not None:
params.extend(["-c", parsed.classindex])
if parsed.graph is not None:
params.extend(["-g", parsed.graph])
jvm.start(jars, max_heap_size=parsed.heap, packages=True)
logger.debug("Commandline: " + join_options(sys.argv[1:]))
try:
clusterer = Clusterer(classname=parsed.clusterer)
if len(parsed.option) > 0:
clusterer.options = parsed.option
print(ClusterEvaluation.evaluate_clusterer(clusterer, params))
except Exception as e:
print(e)
finally:
jvm.stop()
def main(args=None):
"""
Runs attribute selection from the command-line. Calls JVM start/stop automatically.
Use -h to see all options.
:param args: the command-line arguments to use, uses sys.argv if None
:type args: list
"""
parser = argparse.ArgumentParser(
description='Performs attribute selection from the command-line. Calls JVM start/stop automatically.')
parser.add_argument("-j", metavar="classpath", dest="classpath", help="additional classpath, jars/directories")
parser.add_argument("-X", metavar="heap", dest="heap", help="max heap size for jvm, e.g., 512m")
parser.add_argument("-i", metavar="input", dest="input", required=True, help="input file")
parser.add_argument("-c", metavar="class index", dest="classindex", help="1-based class attribute index")
parser.add_argument("-s", metavar="search", dest="search", help="search method, classname and options")
parser.add_argument("-x", metavar="num folds", dest="numfolds", help="number of folds")
parser.add_argument("-n", metavar="seed", dest="seed", help="the seed value for randomization")
parser.add_argument("evaluator", help="evaluator classname, e.g., weka.attributeSelection.CfsSubsetEval")
parser.add_argument("option", nargs=argparse.REMAINDER, help="additional evaluator options")
parsed = parser.parse_args(args=args)
jars = []
if parsed.classpath is not None:
jars = parsed.classpath.split(os.pathsep)
params = []
if parsed.input is not None:
params.extend(["-i", parsed.input])
if parsed.classindex is not None:
params.extend(["-c", parsed.classindex])
if parsed.search is not None:
params.extend(["-s", parsed.search])
if parsed.numfolds is not None:
params.extend(["-x", parsed.numfolds])
if parsed.seed is not None:
params.extend(["-n", parsed.seed])
jvm.start(jars, max_heap_size=parsed.heap, packages=True)
logger.debug("Commandline: " + join_options(sys.argv[1:]))
try:
evaluation = ASEvaluation(classname=parsed.evaluator)
if len(parsed.option) > 0:
evaluation.options = parsed.option
print(AttributeSelection.attribute_selection(evaluation, params))
except Exception as e:
print(traceback.format_exc())
finally:
jvm.stop()
def main():
"""
Runs attribute selection from the command-line. Calls JVM start/stop automatically.
Use -h to see all options.
"""
parser = argparse.ArgumentParser(
description='Performs attribute selection from the command-line. Calls JVM start/stop automatically.')
parser.add_argument("-j", metavar="classpath", dest="classpath", help="additional classpath, jars/directories")
parser.add_argument("-X", metavar="heap", dest="heap", help="max heap size for jvm, e.g., 512m")
parser.add_argument("-i", metavar="input", dest="input", required=True, help="input file")
parser.add_argument("-c", metavar="class index", dest="classindex", help="1-based class attribute index")
parser.add_argument("-s", metavar="search", dest="search", help="search method, classname and options")
parser.add_argument("-x", metavar="num folds", dest="numfolds", help="number of folds")
parser.add_argument("-n", metavar="seed", dest="seed", help="the seed value for randomization")
parser.add_argument("evaluator", help="evaluator classname, e.g., weka.attributeSelection.CfsSubsetEval")
parser.add_argument("option", nargs=argparse.REMAINDER, help="additional evaluator options")
parsed = parser.parse_args()
jars = []
if parsed.classpath is not None:
jars = parsed.classpath.split(os.pathsep)
params = []
if parsed.input is not None:
params.extend(["-i", parsed.input])
if parsed.classindex is not None:
params.extend(["-c", parsed.classindex])
if parsed.search is not None:
params.extend(["-s", parsed.search])
if parsed.numfolds is not None:
params.extend(["-x", parsed.numfolds])
if parsed.seed is not None:
params.extend(["-n", parsed.seed])
jvm.start(jars, max_heap_size=parsed.heap, packages=True)
logger.debug("Commandline: " + join_options(sys.argv[1:]))
try:
evaluation = ASEvaluation(classname=parsed.evaluator)
if len(parsed.option) > 0:
evaluation.options = parsed.option
print(AttributeSelection.attribute_selection(evaluation, params))
except Exception as e:
print(e)
finally:
jvm.stop()
def test_join_options(self):
"""
Tests the join_options method.
"""
self.assertEqual("", str(classes.join_options([])))
self.assertEqual("-t /some/where/test.arff", str(classes.join_options(["-t", "/some/where/test.arff"])))
def plot_learning_curve(classifiers, train, test=None, increments=100, metric="percent_correct",
title="Learning curve", label_template="[#] @ $", key_loc="lower right",
outfile=None, wait=True):
"""
Plots
:param classifiers: list of Classifier template objects
:type classifiers: list of Classifier
:param train: dataset to use for the building the classifier, used for evaluating it test set None
:type train: Instances
:param test: optional dataset to use for the testing the built classifiers
:type test: Instances
:param increments: the increments (>= 1: # of instances, <1: percentage of dataset)
:type increments: float
:param metric: the name of the numeric metric to plot (Evaluation.<metric>)
:type metric: str
:param title: the title for the plot
:type title: str
:param label_template: the template for the label in the plot
(#: 1-based index, @: full classname, !: simple classname, $: options)
:type label_template: str
:param key_loc: the location string for the key
:type key_loc: str
:param outfile: the output file, ignored if None
:type outfile: str
:param wait: whether to wait for the user to close the plot
:type wait: bool
"""
if not plot.matplotlib_available:
logger.error("Matplotlib is not installed, plotting unavailable!")
return
if not train.has_class():
logger.error("Training set has no class attribute set!")
return
if (test is not None) and (train.equal_headers(test) is not None):
logger.error("Training and test set are not compatible: " + train.equal_headers(test))
return
if increments >= 1:
inc = increments
else:
inc = round(train.num_instances * increments)
steps = []
cls = []
evls = {}
for classifier in classifiers:
cl = Classifier.make_copy(classifier)
cls.append(cl)
evls[cl] = []
if test is None:
tst = train
else:
tst = test
for i in xrange(train.num_instances):
if (i > 0) and (i % inc == 0):
steps.append(i+1)
for cl in cls:
# train
if cl.is_updateable:
if i == 0:
tr = Instances.copy_instances(train, 0, 1)
cl.build_classifier(tr)
else:
cl.update_classifier(train.get_instance(i))
else:
if (i > 0) and (i % inc == 0):
tr = Instances.copy_instances(train, 0, i + 1)
cl.build_classifier(tr)
# evaluate
if (i > 0) and (i % inc == 0):
evl = Evaluation(tst)
evl.test_model(cl, tst)
evls[cl].append(getattr(evl, metric))
fig, ax = plt.subplots()
ax.set_xlabel("# of instances")
ax.set_ylabel(metric)
ax.set_title(title)
fig.canvas.set_window_title(title)
ax.grid(True)
i = 0
for cl in cls:
evl = evls[cl]
i += 1
plot_label = label_template.\
replace("#", str(i)).\
replace("@", cl.classname).\
replace("!", cl.classname[cl.classname.rfind(".") + 1:]).\
replace("$", join_options(cl.config))
ax.plot(steps, evl, label=plot_label)
plt.draw()
plt.legend(loc=key_loc, shadow=True)
if outfile is not None:
plt.savefig(outfile)
if wait:
plt.show()
def plot_learning_curve(classifiers,
train,
test=None,
increments=100,
metric="percent_correct",
title="Learning curve",
label_template="[#] @ $",
key_loc="lower right",
outfile=None,
wait=True):
"""
Plots a learning curve.
:param classifiers: list of Classifier template objects
:type classifiers: list of Classifier
:param train: dataset to use for the building the classifier, used for evaluating it test set None
:type train: Instances
:param test: optional dataset (or list of datasets) to use for the testing the built classifiers
:type test: list or Instances
:param increments: the increments (>= 1: # of instances, <1: percentage of dataset)
:type increments: float
:param metric: the name of the numeric metric to plot (Evaluation.<metric>)
:type metric: str
:param title: the title for the plot
:type title: str
:param label_template: the template for the label in the plot
(#: 1-based index of classifier, @: full classname, !: simple classname,
$: options, *: 1-based index of test set)
:type label_template: str
:param key_loc: the location string for the key
:type key_loc: str
:param outfile: the output file, ignored if None
:type outfile: str
:param wait: whether to wait for the user to close the plot
:type wait: bool
"""
if not plot.matplotlib_available:
logger.error("Matplotlib is not installed, plotting unavailable!")
return
if not train.has_class():
logger.error("Training set has no class attribute set!")
return
if increments >= 1:
inc = increments
else:
inc = round(train.num_instances * increments)
if test is None:
tst = [train]
elif isinstance(test, list):
tst = test
elif isinstance(test, Instances):
tst = [test]
else:
logger.error("Expected list or Instances object, instead: " +
type(test))
return
for t in tst:
if train.equal_headers(t) is not None:
logger.error("Training and test set are not compatible: " +
train.equal_headers(t))
return
steps = []
cls = []
evls = {}
for classifier in classifiers:
cl = Classifier.make_copy(classifier)
cls.append(cl)
evls[cl] = {}
for t in tst:
evls[cl][t] = []
for i in xrange(train.num_instances):
if (i > 0) and (i % inc == 0):
steps.append(i + 1)
for cl in cls:
# train
if cl.is_updateable:
if i == 0:
tr = Instances.copy_instances(train, 0, 1)
cl.build_classifier(tr)
else:
cl.update_classifier(train.get_instance(i))
else:
if (i > 0) and (i % inc == 0):
tr = Instances.copy_instances(train, 0, i + 1)
cl.build_classifier(tr)
# evaluate
if (i > 0) and (i % inc == 0):
for t in tst:
evl = Evaluation(t)
evl.test_model(cl, t)
evls[cl][t].append(getattr(evl, metric))
fig, ax = plt.subplots()
ax.set_xlabel("# of instances")
ax.set_ylabel(metric)
ax.set_title(title)
fig.canvas.set_window_title(title)
ax.grid(True)
i = 0
for cl in cls:
evlpertest = evls[cl]
i += 1
n = 0
for t in tst:
evl = evlpertest[t]
n += 1
plot_label = label_template.\
replace("#", str(i)).\
replace("*", str(n)).\
replace("@", cl.classname).\
replace("!", cl.classname[cl.classname.rfind(".") + 1:]).\
replace("$", join_options(cl.config))
ax.plot(steps, evl, label=plot_label)
plt.draw()
plt.legend(loc=key_loc, shadow=True)
if outfile is not None:
plt.savefig(outfile)
if wait:
plt.show()
def main():
"""
Runs a filter from the command-line. Calls JVM start/stop automatically.
Use -h to see all options.
"""
parser = argparse.ArgumentParser(
description=
'Executes a filter from the command-line. Calls JVM start/stop automatically.'
)
parser.add_argument("-j",
metavar="classpath",
dest="classpath",
help="additional classpath, jars/directories")
parser.add_argument("-X",
metavar="heap",
dest="heap",
help="max heap size for jvm, e.g., 512m")
parser.add_argument("-i",
metavar="input1",
dest="input1",
required=True,
help="input file 1")
parser.add_argument("-o",
metavar="output1",
dest="output1",
required=True,
help="output file 1")
parser.add_argument("-r",
metavar="input2",
dest="input2",
help="input file 2")
parser.add_argument("-s",
metavar="output2",
dest="output2",
help="output file 2")
parser.add_argument("-c",
metavar="classindex",
default="-1",
dest="classindex",
help="1-based class attribute index")
parser.add_argument("filter",
help="filter classname, e.g., weka.filters.AllFilter")
parser.add_argument("option",
nargs=argparse.REMAINDER,
help="additional filter options")
parsed = parser.parse_args()
if parsed.input2 is None and parsed.output2 is not None:
raise Exception("No second input file provided ('-r ...')!")
jars = []
if parsed.classpath is not None:
jars = parsed.classpath.split(os.pathsep)
params = []
if parsed.input1 is not None:
params.extend(["-i", parsed.input1])
if parsed.output1 is not None:
params.extend(["-o", parsed.output1])
if parsed.input2 is not None:
params.extend(["-r", parsed.input2])
if parsed.output2 is not None:
params.extend(["-s", parsed.output2])
if parsed.classindex is not None:
params.extend(["-c", parsed.classindex])
jvm.start(jars, max_heap_size=parsed.heap, packages=True)
logger.debug("Commandline: " + join_options(sys.argv[1:]))
try:
flter = Filter(parsed.filter)
if len(parsed.option) > 0:
flter.options = parsed.option
loader = Loader(classname="weka.core.converters.ArffLoader")
in1 = loader.load_file(parsed.input1)
cls = parsed.classindex
if str(parsed.classindex) == "first":
cls = "0"
if str(parsed.classindex) == "last":
cls = str(in1.num_attributes - 1)
in1.class_index = int(cls)
flter.inputformat(in1)
out1 = flter.filter(in1)
saver = Saver(classname="weka.core.converters.ArffSaver")
saver.save_file(out1, parsed.output1)
if parsed.input2 is not None:
in2 = loader.load_file(parsed.input2)
in2.class_index = int(cls)
out2 = flter.filter(in2)
saver.save_file(out2, parsed.output2)
except Exception as e:
print(e)
finally:
jvm.stop()
def main():
"""
Runs a filter from the command-line. Calls JVM start/stop automatically.
Use -h to see all options.
"""
parser = argparse.ArgumentParser(
description='Executes a filter from the command-line. Calls JVM start/stop automatically.')
parser.add_argument("-j", metavar="classpath", dest="classpath", help="additional classpath, jars/directories")
parser.add_argument("-X", metavar="heap", dest="heap", help="max heap size for jvm, e.g., 512m")
parser.add_argument("-i", metavar="input1", dest="input1", required=True, help="input file 1")
parser.add_argument("-o", metavar="output1", dest="output1", required=True, help="output file 1")
parser.add_argument("-r", metavar="input2", dest="input2", help="input file 2")
parser.add_argument("-s", metavar="output2", dest="output2", help="output file 2")
parser.add_argument("-c", metavar="classindex", default="-1", dest="classindex",
help="1-based class attribute index")
parser.add_argument("filter", help="filter classname, e.g., weka.filters.AllFilter")
parser.add_argument("option", nargs=argparse.REMAINDER, help="additional filter options")
parsed = parser.parse_args()
if parsed.input2 is None and parsed.output2 is not None:
raise Exception("No second input file provided ('-r ...')!")
jars = []
if parsed.classpath is not None:
jars = parsed.classpath.split(os.pathsep)
params = []
if parsed.input1 is not None:
params.extend(["-i", parsed.input1])
if parsed.output1 is not None:
params.extend(["-o", parsed.output1])
if parsed.input2 is not None:
params.extend(["-r", parsed.input2])
if parsed.output2 is not None:
params.extend(["-s", parsed.output2])
if parsed.classindex is not None:
params.extend(["-c", parsed.classindex])
jvm.start(jars, max_heap_size=parsed.heap, packages=True)
logger.debug("Commandline: " + join_options(sys.argv[1:]))
try:
flter = Filter(parsed.filter)
if len(parsed.option) > 0:
flter.options = parsed.option
loader = Loader(classname="weka.core.converters.ArffLoader")
in1 = loader.load_file(parsed.input1)
cls = parsed.classindex
if str(parsed.classindex) == "first":
cls = "0"
if str(parsed.classindex) == "last":
cls = str(in1.num_attributes - 1)
in1.class_index = int(cls)
flter.inputformat(in1)
out1 = flter.filter(in1)
saver = Saver(classname="weka.core.converters.ArffSaver")
saver.save_file(out1, parsed.output1)
if parsed.input2 is not None:
in2 = loader.load_file(parsed.input2)
in2.class_index = int(cls)
out2 = flter.filter(in2)
saver.save_file(out2, parsed.output2)
except Exception, e:
print(e)
