Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Enable flake8

Browse Source
This commit is contained in:
sinhrks 2016-04-24 16:34:46 +09:00
parent b3c9e6a0db
commit 8fc2456c87
19 changed files with 282 additions and 199 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
python-package/setup.py
View File

@ -4,7 +4,7 @@ from __future__ import absolute_import
import sys import sys
import os import os
from setuptools import setup, find_packages from setuptools import setup, find_packages
#import subprocess # import subprocess
sys.path.insert(0, '.') sys.path.insert(0, '.')
CURRENT_DIR = os.path.dirname(__file__) CURRENT_DIR = os.path.dirname(__file__)
@ -18,12 +18,12 @@ exec(compile(open(libpath_py, "rb").read(), libpath_py, 'exec'), libpath, libpat
LIB_PATH = libpath['find_lib_path']() LIB_PATH = libpath['find_lib_path']()
print("Install libxgboost from: %s" % LIB_PATH) print("Install libxgboost from: %s" % LIB_PATH)
#Please use setup_pip.py for generating and deploying pip installation # Please use setup_pip.py for generating and deploying pip installation
#detailed instruction in setup_pip.py # detailed instruction in setup_pip.py
setup(name='xgboost', setup(name='xgboost',
version=open(os.path.join(CURRENT_DIR, 'xgboost/VERSION')).read().strip(), version=open(os.path.join(CURRENT_DIR, 'xgboost/VERSION')).read().strip(),
#version='0.4a23', # version='0.4a23',
description = "XGBoost Python Package", description="XGBoost Python Package",
long_description=open(os.path.join(CURRENT_DIR, 'README.rst')).read(), long_description=open(os.path.join(CURRENT_DIR, 'README.rst')).read(),
install_requires=[ install_requires=[
'numpy', 'numpy',
@ -33,8 +33,8 @@ setup(name='xgboost',
maintainer_email='phunter.lau@gmail.com', maintainer_email='phunter.lau@gmail.com',
zip_safe=False, zip_safe=False,
packages=find_packages(), packages=find_packages(),
#this will use MANIFEST.in during install where we specify additional files, # this will use MANIFEST.in during install where we specify additional files,
#this is the golden line # this is the golden line
include_package_data=True, include_package_data=True,
data_files=[('xgboost', LIB_PATH)], data_files=[('xgboost', LIB_PATH)],
url='https://github.com/dmlc/xgboost') url='https://github.com/dmlc/xgboost')

38
python-package/setup_pip.py
View File

@ -4,14 +4,14 @@ from __future__ import absolute_import
import sys import sys
import os import os
from setuptools import setup, find_packages from setuptools import setup, find_packages
#import subprocess # import subprocess
sys.path.insert(0, '.') sys.path.insert(0, '.')
#this script is for packing and shipping pip installation # this script is for packing and shipping pip installation
#it builds xgboost code on the fly and packs for pip # it builds xgboost code on the fly and packs for pip
#please don't use this file for installing from github # please don't use this file for installing from github
if os.name != 'nt': #if not windows, compile and install if os.name != 'nt': # if not windows, compile and install
os.system('sh ./xgboost/build-python.sh') os.system('sh ./xgboost/build-python.sh')
else: else:
print('Windows users please use github installation.') print('Windows users please use github installation.')
@ -28,12 +28,12 @@ exec(compile(open(libpath_py, "rb").read(), libpath_py, 'exec'), libpath, libpat
LIB_PATH = libpath['find_lib_path']() LIB_PATH = libpath['find_lib_path']()
#to deploy to pip, please use # to deploy to pip, please use
#make pythonpack # make pythonpack
#python setup.py register sdist upload # python setup.py register sdist upload
#and be sure to test it firstly using "python setup.py register sdist upload -r pypitest" # and be sure to test it firstly using "python setup.py register sdist upload -r pypitest"
setup(name='xgboost', setup(name='xgboost',
#version=open(os.path.join(CURRENT_DIR, 'xgboost/VERSION')).read().strip(), # version=open(os.path.join(CURRENT_DIR, 'xgboost/VERSION')).read().strip(),
version='0.4a30', version='0.4a30',
description=open(os.path.join(CURRENT_DIR, 'README.rst')).read(), description=open(os.path.join(CURRENT_DIR, 'README.rst')).read(),
install_requires=[ install_requires=[
@ -44,15 +44,15 @@ setup(name='xgboost',
maintainer_email='phunter.lau@gmail.com', maintainer_email='phunter.lau@gmail.com',
zip_safe=False, zip_safe=False,
packages=find_packages(), packages=find_packages(),
#don't need this and don't use this, give everything to MANIFEST.in # don't need this and don't use this, give everything to MANIFEST.in
#package_dir = {'':'xgboost'}, # package_dir = {'':'xgboost'},
#package_data = {'': ['*.txt','*.md','*.sh'], # package_data = {'': ['*.txt','*.md','*.sh'],
# } # }
#this will use MANIFEST.in during install where we specify additional files, # this will use MANIFEST.in during install where we specify additional files,
#this is the golden line # this is the golden line
include_package_data=True, include_package_data=True,
#!!! don't use data_files for creating pip installation, # !!! don't use data_files for creating pip installation,
#otherwise install_data process will copy it to # otherwise install_data process will copy it to
#root directory for some machines, and cause confusions on building # root directory for some machines, and cause confusions on building
#data_files=[('xgboost', LIB_PATH)], # data_files=[('xgboost', LIB_PATH)],
url='https://github.com/dmlc/xgboost') url='https://github.com/dmlc/xgboost')

2
python-package/xgboost/__init__.py
View File

@ -10,7 +10,7 @@ import os
from .core import DMatrix, Booster from .core import DMatrix, Booster
from .training import train, cv from .training import train, cv
from . import rabit from . import rabit # noqa
try: try:
from .sklearn import XGBModel, XGBClassifier, XGBRegressor from .sklearn import XGBModel, XGBClassifier, XGBRegressor
from .plotting import plot_importance, plot_tree, to_graphviz from .plotting import plot_importance, plot_tree, to_graphviz

16
python-package/xgboost/compat.py
View File

@ -12,11 +12,21 @@ PY3 = (sys.version_info[0] == 3)
if PY3: if PY3:
# pylint: disable=invalid-name, redefined-builtin # pylint: disable=invalid-name, redefined-builtin
STRING_TYPES = str, STRING_TYPES = str,
py_str = lambda x: x.decode('utf-8')
def py_str(x):
return x.decode('utf-8')
else: else:
# pylint: disable=invalid-name # pylint: disable=invalid-name
STRING_TYPES = basestring, STRING_TYPES = basestring,
py_str = lambda x: x
def py_str(x):
return x
try:
import cPickle as pickle # noqa
except ImportError:
import pickle # noqa
# pandas # pandas
try: try:
@ -34,7 +44,7 @@ except ImportError:
try: try:
from sklearn.base import BaseEstimator from sklearn.base import BaseEstimator
from sklearn.base import RegressorMixin, ClassifierMixin from sklearn.base import RegressorMixin, ClassifierMixin
from sklearn.preprocessing import LabelEncoder from sklearn.preprocessing import LabelEncoder # noqa
from sklearn.cross_validation import KFold, StratifiedKFold from sklearn.cross_validation import KFold, StratifiedKFold
SKLEARN_INSTALLED = True SKLEARN_INSTALLED = True

22
python-package/xgboost/core.py
View File

@ -14,6 +14,7 @@ from .libpath import find_lib_path
from .compat import STRING_TYPES, PY3, DataFrame, py_str from .compat import STRING_TYPES, PY3, DataFrame, py_str
class XGBoostError(Exception): class XGBoostError(Exception):
"""Error throwed by xgboost trainer.""" """Error throwed by xgboost trainer."""
pass pass
@ -82,6 +83,7 @@ def _load_lib():
# load the XGBoost library globally # load the XGBoost library globally
_LIB = _load_lib() _LIB = _load_lib()
def _check_call(ret): def _check_call(ret):
"""Check the return value of C API call """Check the return value of C API call
@ -129,7 +131,6 @@ def c_array(ctype, values):
return (ctype * len(values))(*values) return (ctype * len(values))(*values)
PANDAS_DTYPE_MAPPER = {'int8': 'int', 'int16': 'int', 'int32': 'int', 'int64': 'int', PANDAS_DTYPE_MAPPER = {'int8': 'int', 'int16': 'int', 'int32': 'int', 'int64': 'int',
'uint8': 'int', 'uint16': 'int', 'uint32': 'int', 'uint64': 'int', 'uint8': 'int', 'uint16': 'int', 'uint32': 'int', 'uint64': 'int',
'float16': 'float', 'float32': 'float', 'float64': 'float', 'float16': 'float', 'float32': 'float', 'float64': 'float',
@ -144,8 +145,12 @@ def _maybe_pandas_data(data, feature_names, feature_types):
data_dtypes = data.dtypes data_dtypes = data.dtypes
if not all(dtype.name in PANDAS_DTYPE_MAPPER for dtype in data_dtypes): if not all(dtype.name in PANDAS_DTYPE_MAPPER for dtype in data_dtypes):
bad_fields = [data.columns[i] for i, dtype in enumerate(data_dtypes) if dtype.name not in PANDAS_DTYPE_MAPPER ] bad_fields = [data.columns[i] for i, dtype in
raise ValueError('DataFrame.dtypes for data must be int, float or bool.\nDid not expect the data types in fie lds '+', '.join(bad_fields)) enumerate(data_dtypes) if dtype.name not in PANDAS_DTYPE_MAPPER]
msg = """DataFrame.dtypes for data must be int, float or bool.
Did not expect the data types in fields """
raise ValueError(msg + ', '.join(bad_fields))
if feature_names is None: if feature_names is None:
feature_names = data.columns.format() feature_names = data.columns.format()
@ -174,6 +179,7 @@ def _maybe_pandas_label(label):
return label return label
class DMatrix(object): class DMatrix(object):
"""Data Matrix used in XGBoost. """Data Matrix used in XGBoost.
@ -1041,8 +1047,14 @@ class Booster(object):
if self.feature_names != data.feature_names: if self.feature_names != data.feature_names:
dat_missing = set(self.feature_names) - set(data.feature_names) dat_missing = set(self.feature_names) - set(data.feature_names)
my_missing = set(data.feature_names) - set(self.feature_names) my_missing = set(data.feature_names) - set(self.feature_names)
msg = 'feature_names mismatch: {0} {1}' msg = 'feature_names mismatch: {0} {1}'
if dat_missing: msg +='\nexpected ' + ', '.join(str(s) for s in dat_missing) +' in input data'
if my_missing: msg +='\ntraining data did not have the following fields: ' + ', '.join(str(s) for s in my_missing) if dat_missing:
msg += '\nexpected ' + ', '.join(str(s) for s in dat_missing) + ' in input data'
if my_missing:
msg += '\ntraining data did not have the following fields: ' + ', '.join(str(s) for s in my_missing)
raise ValueError(msg.format(self.feature_names, raise ValueError(msg.format(self.feature_names,
data.feature_names)) data.feature_names))

3
python-package/xgboost/libpath.py
View File

@ -36,7 +36,8 @@ def find_lib_path():
else: else:
dll_path = [os.path.join(p, 'libxgboost.so') for p in dll_path] dll_path = [os.path.join(p, 'libxgboost.so') for p in dll_path]
lib_path = [p for p in dll_path if os.path.exists(p) and os.path.isfile(p)] lib_path = [p for p in dll_path if os.path.exists(p) and os.path.isfile(p)]
#From github issues, most of installation errors come from machines w/o compilers
# From github issues, most of installation errors come from machines w/o compilers
if len(lib_path) == 0 and not os.environ.get('XGBOOST_BUILD_DOC', False): if len(lib_path) == 0 and not os.environ.get('XGBOOST_BUILD_DOC', False):
raise XGBoostLibraryNotFound( raise XGBoostLibraryNotFound(
'Cannot find XGBoost Libarary in the candicate path, ' + 'Cannot find XGBoost Libarary in the candicate path, ' +

2
python-package/xgboost/plotting.py
View File

@ -10,6 +10,7 @@ import numpy as np
from .core import Booster from .core import Booster
from .sklearn import XGBModel from .sklearn import XGBModel
def plot_importance(booster, ax=None, height=0.2, def plot_importance(booster, ax=None, height=0.2,
xlim=None, ylim=None, title='Feature importance', xlim=None, ylim=None, title='Feature importance',
xlabel='F score', ylabel='Features', xlabel='F score', ylabel='Features',
@ -105,6 +106,7 @@ _LEAFPAT = re.compile(r'(\d+):(leaf=.+)')
_EDGEPAT = re.compile(r'yes=(\d+),no=(\d+),missing=(\d+)') _EDGEPAT = re.compile(r'yes=(\d+),no=(\d+),missing=(\d+)')
_EDGEPAT2 = re.compile(r'yes=(\d+),no=(\d+)') _EDGEPAT2 = re.compile(r'yes=(\d+),no=(\d+)')
def _parse_node(graph, text): def _parse_node(graph, text):
"""parse dumped node""" """parse dumped node"""
match = _NODEPAT.match(text) match = _NODEPAT.match(text)

22
python-package/xgboost/rabit.py
View File

@ -1,11 +1,12 @@
"""Distributed XGBoost Rabit related API.""" """Distributed XGBoost Rabit related API."""
from __future__ import absolute_import from __future__ import absolute_import
import sys import sys
import atexit
import ctypes import ctypes
import numpy as np import numpy as np
from .core import _LIB, c_str, STRING_TYPES from .core import _LIB, c_str, STRING_TYPES
from .compat import pickle
def _init_rabit(): def _init_rabit():
"""internal libary initializer.""" """internal libary initializer."""
@ -15,6 +16,7 @@ def _init_rabit():
_LIB.RabitIsDistributed.restype = ctypes.c_int _LIB.RabitIsDistributed.restype = ctypes.c_int
_LIB.RabitVersionNumber.restype = ctypes.c_int _LIB.RabitVersionNumber.restype = ctypes.c_int
def init(args=None): def init(args=None):
"""Initialize the rabit libary with arguments""" """Initialize the rabit libary with arguments"""
if args is None: if args is None:
@ -73,6 +75,7 @@ def tracker_print(msg):
sys.stdout.write(msg) sys.stdout.write(msg)
sys.stdout.flush() sys.stdout.flush()
def get_processor_name(): def get_processor_name():
"""Get the processor name. """Get the processor name.
@ -127,14 +130,14 @@ def broadcast(data, root):
# enumeration of dtypes # enumeration of dtypes
DTYPE_ENUM__ = { DTYPE_ENUM__ = {
np.dtype('int8') : 0, np.dtype('int8'): 0,
np.dtype('uint8') : 1, np.dtype('uint8'): 1,
np.dtype('int32') : 2, np.dtype('int32'): 2,
np.dtype('uint32') : 3, np.dtype('uint32'): 3,
np.dtype('int64') : 4, np.dtype('int64'): 4,
np.dtype('uint64') : 5, np.dtype('uint64'): 5,
np.dtype('float32') : 6, np.dtype('float32'): 6,
np.dtype('float64') : 7 np.dtype('float64'): 7
} }
@ -175,6 +178,7 @@ def allreduce(data, op, prepare_fun=None):
op, None, None) op, None, None)
else: else:
func_ptr = ctypes.CFUNCTYPE(None, ctypes.c_void_p) func_ptr = ctypes.CFUNCTYPE(None, ctypes.c_void_p)
def pfunc(args): def pfunc(args):
"""prepare function.""" """prepare function."""
prepare_fun(data) prepare_fun(data)

1
python-package/xgboost/sklearn.py
View File

@ -366,7 +366,6 @@ class XGBClassifier(XGBModel, XGBClassifierBase):
self.classes_ = np.unique(y) self.classes_ = np.unique(y)
self.n_classes_ = len(self.classes_) self.n_classes_ = len(self.classes_)
xgb_options = self.get_xgb_params() xgb_options = self.get_xgb_params()
if callable(self.objective): if callable(self.objective):

43
python-package/xgboost/training.py
View File

@ -6,12 +6,12 @@ from __future__ import absolute_import
import sys import sys
import re import re
import os
import numpy as np import numpy as np
from .core import Booster, STRING_TYPES from .core import Booster, STRING_TYPES, XGBoostError
from .compat import (SKLEARN_INSTALLED, XGBStratifiedKFold, XGBKFold) from .compat import (SKLEARN_INSTALLED, XGBStratifiedKFold)
from . import rabit from . import rabit
def train(params, dtrain, num_boost_round=10, evals=(), obj=None, feval=None, def train(params, dtrain, num_boost_round=10, evals=(), obj=None, feval=None,
maximize=False, early_stopping_rounds=None, evals_result=None, maximize=False, early_stopping_rounds=None, evals_result=None,
verbose_eval=True, learning_rates=None, xgb_model=None): verbose_eval=True, learning_rates=None, xgb_model=None):
@ -97,7 +97,7 @@ def train(params, dtrain, num_boost_round=10, evals=(), obj=None, feval=None,
verbose_eval = True if verbose_eval_every_line > 0 else False verbose_eval = True if verbose_eval_every_line > 0 else False
if rabit.get_rank() != 0: if rabit.get_rank() != 0:
verbose_eval = False; verbose_eval = False
if xgb_model is not None: if xgb_model is not None:
if not isinstance(xgb_model, STRING_TYPES): if not isinstance(xgb_model, STRING_TYPES):
@ -135,8 +135,9 @@ def train(params, dtrain, num_boost_round=10, evals=(), obj=None, feval=None,
if isinstance(params, list): if isinstance(params, list):
if len(params) != len(dict(params).items()): if len(params) != len(dict(params).items()):
params = dict(params) params = dict(params)
rabit.tracker_print("Multiple eval metrics have been passed: " \ msg = ("Multiple eval metrics have been passed: "
"'{0}' will be used for early stopping.\n\n".format(params['eval_metric'])) "'{0}' will be used for early stopping.\n\n")
rabit.tracker_print(msg.format(params['eval_metric']))
else: else:
params = dict(params) params = dict(params)
@ -173,7 +174,7 @@ def train(params, dtrain, num_boost_round=10, evals=(), obj=None, feval=None,
# Distributed code: need to resume to this point. # Distributed code: need to resume to this point.
# Skip the first update if it is a recovery step. # Skip the first update if it is a recovery step.
if version % 2 == 0: if version % 2 == 0:
bst.update(dtrain, i, obj) bst.update(dtrain, i, obj)
bst.save_rabit_checkpoint() bst.save_rabit_checkpoint()
version += 1 version += 1
@ -203,7 +204,7 @@ def train(params, dtrain, num_boost_round=10, evals=(), obj=None, feval=None,
evals_idx = evals_name.index(key) evals_idx = evals_name.index(key)
res_per_eval = len(res) // len(evals_name) res_per_eval = len(res) // len(evals_name)
for r in range(res_per_eval): for r in range(res_per_eval):
res_item = res[(evals_idx*res_per_eval) + r] res_item = res[(evals_idx * res_per_eval) + r]
res_key = res_item[0] res_key = res_item[0]
res_val = res_item[1] res_val = res_item[1]
if res_key in evals_result[key]: if res_key in evals_result[key]:
@ -224,7 +225,8 @@ def train(params, dtrain, num_boost_round=10, evals=(), obj=None, feval=None,
elif i - best_iteration >= early_stopping_rounds: elif i - best_iteration >= early_stopping_rounds:
best_msg = bst.attr('best_msg') best_msg = bst.attr('best_msg')
if verbose_eval: if verbose_eval:
rabit.tracker_print("Stopping. Best iteration:\n{}\n\n".format(best_msg)) msg = "Stopping. Best iteration:\n{}\n\n"
rabit.tracker_print(msg.format(best_msg))
break break
# do checkpoint after evaluation, in case evaluation also updates booster. # do checkpoint after evaluation, in case evaluation also updates booster.
bst.save_rabit_checkpoint() bst.save_rabit_checkpoint()
@ -290,6 +292,7 @@ def mknfold(dall, nfold, param, seed, evals=(), fpreproc=None, stratified=False,
ret.append(CVPack(dtrain, dtest, plst)) ret.append(CVPack(dtrain, dtest, plst))
return ret return ret
def aggcv(rlist, show_stdv=True, verbose_eval=None, as_pandas=True, trial=0): def aggcv(rlist, show_stdv=True, verbose_eval=None, as_pandas=True, trial=0):
# pylint: disable=invalid-name # pylint: disable=invalid-name
""" """
@ -405,8 +408,8 @@ def cv(params, dtrain, num_boost_round=10, nfold=3, stratified=False, folds=None
------- -------
evaluation history : list(string) evaluation history : list(string)
""" """
if stratified == True and not SKLEARN_INSTALLED: if stratified is True and not SKLEARN_INSTALLED:
raise XGBoostError('sklearn needs to be installed in order to use stratified cv') raise XGBoostError('sklearn needs to be installed in order to use stratified cv')
if isinstance(metrics, str): if isinstance(metrics, str):
metrics = [metrics] metrics = [metrics]
@ -417,7 +420,7 @@ def cv(params, dtrain, num_boost_round=10, nfold=3, stratified=False, folds=None
if 'eval_metric' in params: if 'eval_metric' in params:
params['eval_metric'] = _metrics params['eval_metric'] = _metrics
else: else:
params= dict((k, v) for k, v in params.items()) params = dict((k, v) for k, v in params.items())
if len(metrics) == 0 and 'eval_metric' in params: if len(metrics) == 0 and 'eval_metric' in params:
if isinstance(params['eval_metric'], list): if isinstance(params['eval_metric'], list):
@ -428,12 +431,14 @@ def cv(params, dtrain, num_boost_round=10, nfold=3, stratified=False, folds=None
params.pop("eval_metric", None) params.pop("eval_metric", None)
if early_stopping_rounds is not None: if early_stopping_rounds is not None:
if len(metrics) > 1: if len(metrics) > 1:
raise ValueError('Check your params. '\ msg = ('Check your params. '
'Early stopping works with single eval metric only.') 'Early stopping works with single eval metric only.')
raise ValueError(msg)
if verbose_eval: if verbose_eval:
sys.stderr.write("Will train until cv error hasn't decreased in {} rounds.\n".format(\ msg = "Will train until cv error hasn't decreased in {} rounds.\n"
early_stopping_rounds)) sys.stderr.write(msg.format(early_stopping_rounds))
maximize_score = False maximize_score = False
if len(metrics) == 1: if len(metrics) == 1:
@ -466,10 +471,10 @@ def cv(params, dtrain, num_boost_round=10, nfold=3, stratified=False, folds=None
best_score = score best_score = score
best_score_i = i best_score_i = i
elif i - best_score_i >= early_stopping_rounds: elif i - best_score_i >= early_stopping_rounds:
results = results[:best_score_i+1] results = results[:best_score_i + 1]
if verbose_eval: if verbose_eval:
sys.stderr.write("Stopping. Best iteration:\n[{}] cv-mean:{}\tcv-std:{}\n". msg = "Stopping. Best iteration:\n[{}] cv-mean:{}\tcv-std:{}\n"
format(best_score_i, results[-1][0], results[-1][1])) sys.stderr.write(msg.format(best_score_i, results[-1][0], results[-1][1]))
break break
if as_pandas: if as_pandas:
try: try:

4
tests/python/test_basic.py
View File

@ -8,6 +8,7 @@ rng = np.random.RandomState(1994)
class TestBasic(unittest.TestCase): class TestBasic(unittest.TestCase):
def test_basic(self): def test_basic(self):
dtrain = xgb.DMatrix(dpath + 'agaricus.txt.train') dtrain = xgb.DMatrix(dpath + 'agaricus.txt.train')
dtest = xgb.DMatrix(dpath + 'agaricus.txt.test') dtest = xgb.DMatrix(dpath + 'agaricus.txt.test')
@ -37,7 +38,7 @@ class TestBasic(unittest.TestCase):
def test_multiclass(self): def test_multiclass(self):
dtrain = xgb.DMatrix(dpath + 'agaricus.txt.train') dtrain = xgb.DMatrix(dpath + 'agaricus.txt.train')
dtest = xgb.DMatrix(dpath + 'agaricus.txt.test') dtest = xgb.DMatrix(dpath + 'agaricus.txt.test')
param = {'max_depth': 2, 'eta': 1, 'silent': 1, 'num_class' : 2} param = {'max_depth': 2, 'eta': 1, 'silent': 1, 'num_class': 2}
# specify validations set to watch performance # specify validations set to watch performance
watchlist = [(dtest, 'eval'), (dtrain, 'train')] watchlist = [(dtest, 'eval'), (dtrain, 'train')]
num_round = 2 num_round = 2
@ -60,7 +61,6 @@ class TestBasic(unittest.TestCase):
# assert they are the same # assert they are the same
assert np.sum(np.abs(preds2 - preds)) == 0 assert np.sum(np.abs(preds2 - preds)) == 0
def test_dmatrix_init(self): def test_dmatrix_init(self):
data = np.random.randn(5, 5) data = np.random.randn(5, 5)

148
tests/python/test_basic_models.py
View File

@ -8,82 +8,94 @@ dtest = xgb.DMatrix(dpath + 'agaricus.txt.test')
rng = np.random.RandomState(1994) rng = np.random.RandomState(1994)
class TestModels(unittest.TestCase): class TestModels(unittest.TestCase):
def test_glm(self): def test_glm(self):
param = {'silent':1, 'objective':'binary:logistic', 'booster':'gblinear', 'alpha': 0.0001, 'lambda': 1 } param = {'silent': 1, 'objective': 'binary:logistic',
watchlist = [(dtest,'eval'), (dtrain,'train')] 'booster': 'gblinear', 'alpha': 0.0001, 'lambda': 1}
num_round = 4 watchlist = [(dtest, 'eval'), (dtrain, 'train')]
bst = xgb.train(param, dtrain, num_round, watchlist) num_round = 4
assert isinstance(bst, xgb.core.Booster) bst = xgb.train(param, dtrain, num_round, watchlist)
preds = bst.predict(dtest) assert isinstance(bst, xgb.core.Booster)
labels = dtest.get_label() preds = bst.predict(dtest)
err = sum(1 for i in range(len(preds)) if int(preds[i]>0.5)!=labels[i]) / float(len(preds)) labels = dtest.get_label()
assert err < 0.1 err = sum(1 for i in range(len(preds))
if int(preds[i] > 0.5) != labels[i]) / float(len(preds))
assert err < 0.1
def test_eta_decay(self): def test_eta_decay(self):
param = {'max_depth':2, 'eta':1, 'silent':1, 'objective':'binary:logistic' } param = {'max_depth': 2, 'eta': 1, 'silent': 1, 'objective': 'binary:logistic'}
watchlist = [(dtest,'eval'), (dtrain,'train')] watchlist = [(dtest, 'eval'), (dtrain, 'train')]
num_round = 2 num_round = 2
# learning_rates as a list # learning_rates as a list
bst = xgb.train(param, dtrain, num_round, watchlist, learning_rates=[0.4, 0.3]) bst = xgb.train(param, dtrain, num_round, watchlist, learning_rates=[0.4, 0.3])
assert isinstance(bst, xgb.core.Booster) assert isinstance(bst, xgb.core.Booster)
# learning_rates as a customized decay function # learning_rates as a customized decay function
def eta_decay(ithround, num_boost_round): def eta_decay(ithround, num_boost_round):
return num_boost_round / (ithround + 1) return num_boost_round / (ithround + 1)
bst = xgb.train(param, dtrain, num_round, watchlist, learning_rates=eta_decay)
assert isinstance(bst, xgb.core.Booster)
bst = xgb.train(param, dtrain, num_round, watchlist, learning_rates=eta_decay)
assert isinstance(bst, xgb.core.Booster)
def test_custom_objective(self): def test_custom_objective(self):
param = {'max_depth':2, 'eta':1, 'silent':1 } param = {'max_depth': 2, 'eta': 1, 'silent': 1}
watchlist = [(dtest,'eval'), (dtrain,'train')] watchlist = [(dtest, 'eval'), (dtrain, 'train')]
num_round = 2 num_round = 2
def logregobj(preds, dtrain):
labels = dtrain.get_label()
preds = 1.0 / (1.0 + np.exp(-preds))
grad = preds - labels
hess = preds * (1.0-preds)
return grad, hess
def evalerror(preds, dtrain):
labels = dtrain.get_label()
return 'error', float(sum(labels != (preds > 0.0))) / len(labels)
# test custom_objective in training def logregobj(preds, dtrain):
bst = xgb.train(param, dtrain, num_round, watchlist, logregobj, evalerror) labels = dtrain.get_label()
assert isinstance(bst, xgb.core.Booster) preds = 1.0 / (1.0 + np.exp(-preds))
preds = bst.predict(dtest) grad = preds - labels
labels = dtest.get_label() hess = preds * (1.0 - preds)
err = sum(1 for i in range(len(preds)) if int(preds[i]>0.5)!=labels[i]) / float(len(preds)) return grad, hess
assert err < 0.1
# test custom_objective in cross-validation def evalerror(preds, dtrain):
xgb.cv(param, dtrain, num_round, nfold = 5, seed = 0, labels = dtrain.get_label()
obj = logregobj, feval=evalerror) return 'error', float(sum(labels != (preds > 0.0))) / len(labels)
# test maximize parameter # test custom_objective in training
def neg_evalerror(preds, dtrain): bst = xgb.train(param, dtrain, num_round, watchlist, logregobj, evalerror)
labels = dtrain.get_label() assert isinstance(bst, xgb.core.Booster)
return 'error', float(sum(labels == (preds > 0.0))) / len(labels) preds = bst.predict(dtest)
bst2 = xgb.train(param, dtrain, num_round, watchlist, logregobj, neg_evalerror, maximize=True) labels = dtest.get_label()
preds2 = bst2.predict(dtest) err = sum(1 for i in range(len(preds))
err2 = sum(1 for i in range(len(preds2)) if int(preds2[i]>0.5)!=labels[i]) / float(len(preds2)) if int(preds[i] > 0.5) != labels[i]) / float(len(preds))
assert err == err2 assert err < 0.1
def test_fpreproc(self): # test custom_objective in cross-validation
param = {'max_depth':2, 'eta':1, 'silent':1, 'objective':'binary:logistic'} xgb.cv(param, dtrain, num_round, nfold=5, seed=0,
num_round = 2 obj=logregobj, feval=evalerror)
def fpreproc(dtrain, dtest, param):
label = dtrain.get_label()
ratio = float(np.sum(label == 0)) / np.sum(label==1)
param['scale_pos_weight'] = ratio
return (dtrain, dtest, param)
xgb.cv(param, dtrain, num_round, nfold=5,
metrics={'auc'}, seed = 0, fpreproc = fpreproc)
def test_show_stdv(self): # test maximize parameter
param = {'max_depth':2, 'eta':1, 'silent':1, 'objective':'binary:logistic'} def neg_evalerror(preds, dtrain):
num_round = 2 labels = dtrain.get_label()
xgb.cv(param, dtrain, num_round, nfold=5, return 'error', float(sum(labels == (preds > 0.0))) / len(labels)
metrics={'error'}, seed = 0, show_stdv = False)
bst2 = xgb.train(param, dtrain, num_round, watchlist, logregobj, neg_evalerror, maximize=True)
preds2 = bst2.predict(dtest)
err2 = sum(1 for i in range(len(preds2))
if int(preds2[i] > 0.5) != labels[i]) / float(len(preds2))
assert err == err2
def test_fpreproc(self):
param = {'max_depth': 2, 'eta': 1, 'silent': 1,
'objective': 'binary:logistic'}
num_round = 2
def fpreproc(dtrain, dtest, param):
label = dtrain.get_label()
ratio = float(np.sum(label == 0)) / np.sum(label == 1)
param['scale_pos_weight'] = ratio
return (dtrain, dtest, param)
xgb.cv(param, dtrain, num_round, nfold=5,
metrics={'auc'}, seed=0, fpreproc=fpreproc)
def test_show_stdv(self):
param = {'max_depth': 2, 'eta': 1, 'silent': 1,
'objective': 'binary:logistic'}
num_round = 2
xgb.cv(param, dtrain, num_round, nfold=5,
metrics={'error'}, seed=0, show_stdv=False)

3
tests/python/test_early_stopping.py
View File

@ -1,7 +1,7 @@
import xgboost as xgb import xgboost as xgb
import numpy as np import numpy as np
from sklearn.datasets import load_digits from sklearn.datasets import load_digits
from sklearn.cross_validation import KFold, train_test_split from sklearn.cross_validation import train_test_split
from sklearn.metrics import mean_squared_error from sklearn.metrics import mean_squared_error
import unittest import unittest
@ -40,7 +40,6 @@ class TestEarlyStopping(unittest.TestCase):
dm = xgb.DMatrix(X, label=y) dm = xgb.DMatrix(X, label=y)
params = {'max_depth': 2, 'eta': 1, 'silent': 1, 'objective': 'binary:logistic'} params = {'max_depth': 2, 'eta': 1, 'silent': 1, 'objective': 'binary:logistic'}
import pandas as pd
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, early_stopping_rounds=10) cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, early_stopping_rounds=10)
assert cv.shape[0] == 10 assert cv.shape[0] == 10
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, early_stopping_rounds=5) cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, early_stopping_rounds=5)

5
tests/python/test_eval_metrics.py
View File

@ -1,9 +1,8 @@
import xgboost as xgb import xgboost as xgb
import numpy as np import numpy as np
from sklearn.cross_validation import KFold, train_test_split from sklearn.cross_validation import train_test_split
from sklearn.metrics import mean_squared_error from sklearn.metrics import mean_squared_error
from sklearn.grid_search import GridSearchCV from sklearn.datasets import load_digits
from sklearn.datasets import load_iris, load_digits, load_boston
import unittest import unittest
rng = np.random.RandomState(1337) rng = np.random.RandomState(1337)

1
tests/python/test_plotting.py
View File

@ -12,6 +12,7 @@ matplotlib.use('Agg')
dpath = 'demo/data/' dpath = 'demo/data/'
rng = np.random.RandomState(1994) rng = np.random.RandomState(1994)
class TestPlotting(unittest.TestCase): class TestPlotting(unittest.TestCase):
def test_plotting(self): def test_plotting(self):
bst2 = xgb.Booster(model_file='xgb.model') bst2 = xgb.Booster(model_file='xgb.model')

41
tests/python/test_training_continuation.py
View File

@ -1,10 +1,7 @@
import xgboost as xgb import xgboost as xgb
import numpy as np import numpy as np
from sklearn.preprocessing import MultiLabelBinarizer
from sklearn.cross_validation import KFold, train_test_split
from sklearn.metrics import mean_squared_error from sklearn.metrics import mean_squared_error
from sklearn.grid_search import GridSearchCV from sklearn.datasets import load_digits
from sklearn.datasets import load_iris, load_digits, load_boston
import unittest import unittest
rng = np.random.RandomState(1337) rng = np.random.RandomState(1337)
@ -57,10 +54,14 @@ class TestTrainingContinuation(unittest.TestCase):
ntrees_02b = len(gbdt_02b.get_dump()) ntrees_02b = len(gbdt_02b.get_dump())
assert ntrees_02a == 10 assert ntrees_02a == 10
assert ntrees_02b == 10 assert ntrees_02b == 10
assert mean_squared_error(y_2class, gbdt_01.predict(dtrain_2class)) == \
mean_squared_error(y_2class, gbdt_02a.predict(dtrain_2class)) res1 = mean_squared_error(y_2class, gbdt_01.predict(dtrain_2class))
assert mean_squared_error(y_2class, gbdt_01.predict(dtrain_2class)) == \ res2 = mean_squared_error(y_2class, gbdt_02a.predict(dtrain_2class))
mean_squared_error(y_2class, gbdt_02b.predict(dtrain_2class)) assert res1 == res2
res1 = mean_squared_error(y_2class, gbdt_01.predict(dtrain_2class))
res2 = mean_squared_error(y_2class, gbdt_02b.predict(dtrain_2class))
assert res1 == res2
gbdt_03 = xgb.train(self.xgb_params_01, dtrain_2class, num_boost_round=3) gbdt_03 = xgb.train(self.xgb_params_01, dtrain_2class, num_boost_round=3)
gbdt_03.save_model('xgb_tc.model') gbdt_03.save_model('xgb_tc.model')
@ -71,22 +72,30 @@ class TestTrainingContinuation(unittest.TestCase):
ntrees_03b = len(gbdt_03b.get_dump()) ntrees_03b = len(gbdt_03b.get_dump())
assert ntrees_03a == 10 assert ntrees_03a == 10
assert ntrees_03b == 10 assert ntrees_03b == 10
assert mean_squared_error(y_2class, gbdt_03a.predict(dtrain_2class)) == \
mean_squared_error(y_2class, gbdt_03b.predict(dtrain_2class)) res1 = mean_squared_error(y_2class, gbdt_03a.predict(dtrain_2class))
res2 = mean_squared_error(y_2class, gbdt_03b.predict(dtrain_2class))
assert res1 == res2
gbdt_04 = xgb.train(self.xgb_params_02, dtrain_2class, num_boost_round=3) gbdt_04 = xgb.train(self.xgb_params_02, dtrain_2class, num_boost_round=3)
assert gbdt_04.best_ntree_limit == (gbdt_04.best_iteration + 1) * self.num_parallel_tree assert gbdt_04.best_ntree_limit == (gbdt_04.best_iteration + 1) * self.num_parallel_tree
assert mean_squared_error(y_2class, gbdt_04.predict(dtrain_2class)) == \
mean_squared_error(y_2class, gbdt_04.predict(dtrain_2class, ntree_limit=gbdt_04.best_ntree_limit)) res1 = mean_squared_error(y_2class, gbdt_04.predict(dtrain_2class))
res2 = mean_squared_error(y_2class, gbdt_04.predict(dtrain_2class, ntree_limit=gbdt_04.best_ntree_limit))
assert res1 == res2
gbdt_04 = xgb.train(self.xgb_params_02, dtrain_2class, num_boost_round=7, xgb_model=gbdt_04) gbdt_04 = xgb.train(self.xgb_params_02, dtrain_2class, num_boost_round=7, xgb_model=gbdt_04)
assert gbdt_04.best_ntree_limit == (gbdt_04.best_iteration + 1) * self.num_parallel_tree assert gbdt_04.best_ntree_limit == (gbdt_04.best_iteration + 1) * self.num_parallel_tree
assert mean_squared_error(y_2class, gbdt_04.predict(dtrain_2class)) == \
mean_squared_error(y_2class, gbdt_04.predict(dtrain_2class, ntree_limit=gbdt_04.best_ntree_limit)) res1 = mean_squared_error(y_2class, gbdt_04.predict(dtrain_2class))
res2 = mean_squared_error(y_2class, gbdt_04.predict(dtrain_2class, ntree_limit=gbdt_04.best_ntree_limit))
assert res1 == res2
gbdt_05 = xgb.train(self.xgb_params_03, dtrain_5class, num_boost_round=7) gbdt_05 = xgb.train(self.xgb_params_03, dtrain_5class, num_boost_round=7)
assert gbdt_05.best_ntree_limit == (gbdt_05.best_iteration + 1) * self.num_parallel_tree assert gbdt_05.best_ntree_limit == (gbdt_05.best_iteration + 1) * self.num_parallel_tree
gbdt_05 = xgb.train(self.xgb_params_03, dtrain_5class, num_boost_round=3, xgb_model=gbdt_05) gbdt_05 = xgb.train(self.xgb_params_03, dtrain_5class, num_boost_round=3, xgb_model=gbdt_05)
assert gbdt_05.best_ntree_limit == (gbdt_05.best_iteration + 1) * self.num_parallel_tree assert gbdt_05.best_ntree_limit == (gbdt_05.best_iteration + 1) * self.num_parallel_tree
assert np.any(gbdt_05.predict(dtrain_5class) !=
gbdt_05.predict(dtrain_5class, ntree_limit=gbdt_05.best_ntree_limit)) == False res1 = gbdt_05.predict(dtrain_5class)
res2 = gbdt_05.predict(dtrain_5class, ntree_limit=gbdt_05.best_ntree_limit)
np.testing.assert_almost_equal(res1, res2)

38
tests/python/test_with_pandas.py
View File

@ -111,43 +111,55 @@ class TestPandas(unittest.TestCase):
u'train-error-mean', u'train-error-std']) u'train-error-mean', u'train-error-std'])
assert cv.columns.equals(exp) assert cv.columns.equals(exp)
params = {'max_depth': 2, 'eta': 1, 'silent': 1, 'objective': 'binary:logistic', 'eval_metric': 'auc'} params = {'max_depth': 2, 'eta': 1, 'silent': 1,
'objective': 'binary:logistic', 'eval_metric': 'auc'}
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, as_pandas=True) cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, as_pandas=True)
assert 'eval_metric' in params assert 'eval_metric' in params
assert 'auc' in cv.columns[0] assert 'auc' in cv.columns[0]
params = {'max_depth': 2, 'eta': 1, 'silent': 1, 'objective': 'binary:logistic', 'eval_metric': ['auc']} params = {'max_depth': 2, 'eta': 1, 'silent': 1,
'objective': 'binary:logistic', 'eval_metric': ['auc']}
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, as_pandas=True) cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, as_pandas=True)
assert 'eval_metric' in params assert 'eval_metric' in params
assert 'auc' in cv.columns[0] assert 'auc' in cv.columns[0]
params = {'max_depth': 2, 'eta': 1, 'silent': 1, 'objective': 'binary:logistic', 'eval_metric': ['auc']} params = {'max_depth': 2, 'eta': 1, 'silent': 1,
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, as_pandas=True, early_stopping_rounds=1) 'objective': 'binary:logistic', 'eval_metric': ['auc']}
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10,
as_pandas=True, early_stopping_rounds=1)
assert 'eval_metric' in params assert 'eval_metric' in params
assert 'auc' in cv.columns[0] assert 'auc' in cv.columns[0]
assert cv.shape[0] < 10 assert cv.shape[0] < 10
params = {'max_depth': 2, 'eta': 1, 'silent': 1, 'objective': 'binary:logistic'} params = {'max_depth': 2, 'eta': 1, 'silent': 1,
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, as_pandas=True, metrics='auc') 'objective': 'binary:logistic'}
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10,
as_pandas=True, metrics='auc')
assert 'auc' in cv.columns[0] assert 'auc' in cv.columns[0]
params = {'max_depth': 2, 'eta': 1, 'silent': 1, 'objective': 'binary:logistic'} params = {'max_depth': 2, 'eta': 1, 'silent': 1,
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, as_pandas=True, metrics=['auc']) 'objective': 'binary:logistic'}
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10,
as_pandas=True, metrics=['auc'])
assert 'auc' in cv.columns[0] assert 'auc' in cv.columns[0]
params = {'max_depth': 2, 'eta': 1, 'silent': 1, 'objective': 'binary:logistic', 'eval_metric': ['auc']} params = {'max_depth': 2, 'eta': 1, 'silent': 1,
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, as_pandas=True, metrics='error') 'objective': 'binary:logistic', 'eval_metric': ['auc']}
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10,
as_pandas=True, metrics='error')
assert 'eval_metric' in params assert 'eval_metric' in params
assert 'auc' not in cv.columns[0] assert 'auc' not in cv.columns[0]
assert 'error' in cv.columns[0] assert 'error' in cv.columns[0]
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, as_pandas=True, metrics=['error']) cv = xgb.cv(params, dm, num_boost_round=10, nfold=10,
as_pandas=True, metrics=['error'])
assert 'eval_metric' in params assert 'eval_metric' in params
assert 'auc' not in cv.columns[0] assert 'auc' not in cv.columns[0]
assert 'error' in cv.columns[0] assert 'error' in cv.columns[0]
params = list(params.items()) params = list(params.items())
cv = xgb.cv(params, dm, num_boost_round=10, nfold=10, as_pandas=True, metrics=['error']) cv = xgb.cv(params, dm, num_boost_round=10, nfold=10,
as_pandas=True, metrics=['error'])
assert isinstance(params, list) assert isinstance(params, list)
assert 'auc' not in cv.columns[0] assert 'auc' not in cv.columns[0]
assert 'error' in cv.columns[0] assert 'error' in cv.columns[0]

75
tests/python/test_with_sklearn.py
View File

@ -1,6 +1,5 @@
import xgboost as xgb import xgboost as xgb
import numpy as np import numpy as np
from sklearn.cross_validation import KFold
from sklearn.metrics import mean_squared_error from sklearn.metrics import mean_squared_error
from sklearn.grid_search import GridSearchCV from sklearn.grid_search import GridSearchCV
from sklearn.datasets import load_iris, load_digits, load_boston from sklearn.datasets import load_iris, load_digits, load_boston
@ -8,33 +7,46 @@ from sklearn.cross_validation import KFold, StratifiedKFold, train_test_split
rng = np.random.RandomState(1994) rng = np.random.RandomState(1994)
def test_binary_classification(): def test_binary_classification():
digits = load_digits(2) digits = load_digits(2)
y = digits['target'] y = digits['target']
X = digits['data'] X = digits['data']
kf = KFold(y.shape[0], n_folds=2, shuffle=True, random_state=rng) kf = KFold(y.shape[0], n_folds=2, shuffle=True, random_state=rng)
for train_index, test_index in kf: for train_index, test_index in kf:
xgb_model = xgb.XGBClassifier().fit(X[train_index],y[train_index]) xgb_model = xgb.XGBClassifier().fit(X[train_index], y[train_index])
preds = xgb_model.predict(X[test_index]) preds = xgb_model.predict(X[test_index])
labels = y[test_index] labels = y[test_index]
err = sum(1 for i in range(len(preds)) if int(preds[i]>0.5)!=labels[i]) / float(len(preds)) err = sum(1 for i in range(len(preds))
assert err < 0.1 if int(preds[i] > 0.5) != labels[i]) / float(len(preds))
assert err < 0.1
def test_multiclass_classification(): def test_multiclass_classification():
def check_pred(preds, labels):
err = sum(1 for i in range(len(preds))
if int(preds[i] > 0.5) != labels[i]) / float(len(preds))
assert err < 0.4
iris = load_iris() iris = load_iris()
y = iris['target'] y = iris['target']
X = iris['data'] X = iris['data']
kf = KFold(y.shape[0], n_folds=2, shuffle=True, random_state=rng) kf = KFold(y.shape[0], n_folds=2, shuffle=True, random_state=rng)
for train_index, test_index in kf: for train_index, test_index in kf:
xgb_model = xgb.XGBClassifier().fit(X[train_index],y[train_index]) xgb_model = xgb.XGBClassifier().fit(X[train_index], y[train_index])
preds = xgb_model.predict(X[test_index]) preds = xgb_model.predict(X[test_index])
# test other params in XGBClassifier().fit # test other params in XGBClassifier().fit
preds2 = xgb_model.predict(X[test_index], output_margin=True, ntree_limit=3) preds2 = xgb_model.predict(X[test_index], output_margin=True, ntree_limit=3)
preds3 = xgb_model.predict(X[test_index], output_margin=True, ntree_limit=0) preds3 = xgb_model.predict(X[test_index], output_margin=True, ntree_limit=0)
preds4 = xgb_model.predict(X[test_index], output_margin=False, ntree_limit=3) preds4 = xgb_model.predict(X[test_index], output_margin=False, ntree_limit=3)
labels = y[test_index] labels = y[test_index]
err = sum(1 for i in range(len(preds)) if int(preds[i]>0.5)!=labels[i]) / float(len(preds))
assert err < 0.4 check_pred(preds, labels)
check_pred(preds2, labels)
check_pred(preds3, labels)
check_pred(preds4, labels)
def test_boston_housing_regression(): def test_boston_housing_regression():
boston = load_boston() boston = load_boston()
@ -42,27 +54,33 @@ def test_boston_housing_regression():
X = boston['data'] X = boston['data']
kf = KFold(y.shape[0], n_folds=2, shuffle=True, random_state=rng) kf = KFold(y.shape[0], n_folds=2, shuffle=True, random_state=rng)
for train_index, test_index in kf: for train_index, test_index in kf:
xgb_model = xgb.XGBRegressor().fit(X[train_index],y[train_index]) xgb_model = xgb.XGBRegressor().fit(X[train_index], y[train_index])
preds = xgb_model.predict(X[test_index]) preds = xgb_model.predict(X[test_index])
# test other params in XGBRegressor().fit # test other params in XGBRegressor().fit
preds2 = xgb_model.predict(X[test_index], output_margin=True, ntree_limit=3) preds2 = xgb_model.predict(X[test_index], output_margin=True, ntree_limit=3)
preds3 = xgb_model.predict(X[test_index], output_margin=True, ntree_limit=0) preds3 = xgb_model.predict(X[test_index], output_margin=True, ntree_limit=0)
preds4 = xgb_model.predict(X[test_index], output_margin=False, ntree_limit=3) preds4 = xgb_model.predict(X[test_index], output_margin=False, ntree_limit=3)
labels = y[test_index] labels = y[test_index]
assert mean_squared_error(preds, labels) < 25
assert mean_squared_error(preds, labels) < 25
assert mean_squared_error(preds2, labels) < 350
assert mean_squared_error(preds3, labels) < 25
assert mean_squared_error(preds4, labels) < 350
def test_parameter_tuning(): def test_parameter_tuning():
boston = load_boston() boston = load_boston()
y = boston['target'] y = boston['target']
X = boston['data'] X = boston['data']
xgb_model = xgb.XGBRegressor() xgb_model = xgb.XGBRegressor()
clf = GridSearchCV(xgb_model, clf = GridSearchCV(xgb_model, {'max_depth': [2, 4, 6],
{'max_depth': [2,4,6], 'n_estimators': [50, 100, 200]}, verbose=1)
'n_estimators': [50,100,200]}, verbose=1) clf.fit(X, y)
clf.fit(X,y)
assert clf.best_score_ < 0.7 assert clf.best_score_ < 0.7
assert clf.best_params_ == {'n_estimators': 100, 'max_depth': 4} assert clf.best_params_ == {'n_estimators': 100, 'max_depth': 4}
def test_regression_with_custom_objective(): def test_regression_with_custom_objective():
def objective_ls(y_true, y_pred): def objective_ls(y_true, y_pred):
grad = (y_pred - y_true) grad = (y_pred - y_true)
@ -86,20 +104,17 @@ def test_regression_with_custom_objective():
pass pass
def dummy_objective(y_true, y_pred): def dummy_objective(y_true, y_pred):
raise XGBCustomObjectiveException() raise XGBCustomObjectiveException()
xgb_model = xgb.XGBRegressor(objective=dummy_objective) xgb_model = xgb.XGBRegressor(objective=dummy_objective)
np.testing.assert_raises( np.testing.assert_raises(XGBCustomObjectiveException, xgb_model.fit, X, y)
XGBCustomObjectiveException,
xgb_model.fit,
X, y
)
def test_classification_with_custom_objective(): def test_classification_with_custom_objective():
def logregobj(y_true, y_pred): def logregobj(y_true, y_pred):
y_pred = 1.0 / (1.0 + np.exp(-y_pred)) y_pred = 1.0 / (1.0 + np.exp(-y_pred))
grad = y_pred - y_true grad = y_pred - y_true
hess = y_pred * (1.0-y_pred) hess = y_pred * (1.0 - y_pred)
return grad, hess return grad, hess
digits = load_digits(2) digits = load_digits(2)
@ -107,22 +122,20 @@ def test_classification_with_custom_objective():
X = digits['data'] X = digits['data']
kf = KFold(y.shape[0], n_folds=2, shuffle=True, random_state=rng) kf = KFold(y.shape[0], n_folds=2, shuffle=True, random_state=rng)
for train_index, test_index in kf: for train_index, test_index in kf:
xgb_model = xgb.XGBClassifier(objective=logregobj).fit( xgb_model = xgb.XGBClassifier(objective=logregobj)
X[train_index],y[train_index] xgb_model.fit(X[train_index], y[train_index])
)
preds = xgb_model.predict(X[test_index]) preds = xgb_model.predict(X[test_index])
labels = y[test_index] labels = y[test_index]
err = sum(1 for i in range(len(preds)) err = sum(1 for i in range(len(preds))
if int(preds[i]>0.5)!=labels[i]) / float(len(preds)) if int(preds[i] > 0.5) != labels[i]) / float(len(preds))
assert err < 0.1 assert err < 0.1
# Test that the custom objective function is actually used # Test that the custom objective function is actually used
class XGBCustomObjectiveException(Exception): class XGBCustomObjectiveException(Exception):
pass pass
def dummy_objective(y_true, y_preds): def dummy_objective(y_true, y_preds):
raise XGBCustomObjectiveException() raise XGBCustomObjectiveException()
xgb_model = xgb.XGBClassifier(objective=dummy_objective) xgb_model = xgb.XGBClassifier(objective=dummy_objective)
np.testing.assert_raises( np.testing.assert_raises(
@ -131,6 +144,7 @@ def test_classification_with_custom_objective():
X, y X, y
) )
def test_sklearn_api(): def test_sklearn_api():
iris = load_iris() iris = load_iris()
tr_d, te_d, tr_l, te_l = train_test_split(iris.data, iris.target, train_size=120) tr_d, te_d, tr_l, te_l = train_test_split(iris.data, iris.target, train_size=120)
@ -143,6 +157,7 @@ def test_sklearn_api():
err = sum([1 for p, l in zip(preds, labels) if p != l]) / len(te_l) err = sum([1 for p, l in zip(preds, labels) if p != l]) / len(te_l)
assert err < 0.2 assert err < 0.2
def test_sklearn_plotting(): def test_sklearn_plotting():
iris = load_iris() iris = load_iris()
@ -168,12 +183,13 @@ def test_sklearn_plotting():
ax = xgb.plot_tree(classifier, num_trees=0) ax = xgb.plot_tree(classifier, num_trees=0)
assert isinstance(ax, Axes) assert isinstance(ax, Axes)
def test_sklearn_nfolds_cv(): def test_sklearn_nfolds_cv():
digits = load_digits(3) digits = load_digits(3)
X = digits['data'] X = digits['data']
y = digits['target'] y = digits['target']
dm = xgb.DMatrix(X, label=y) dm = xgb.DMatrix(X, label=y)
params = { params = {
'max_depth': 2, 'max_depth': 2,
'eta': 1, 'eta': 1,
@ -187,9 +203,8 @@ def test_sklearn_nfolds_cv():
nfolds = 5 nfolds = 5
skf = StratifiedKFold(y, n_folds=nfolds, shuffle=True, random_state=seed) skf = StratifiedKFold(y, n_folds=nfolds, shuffle=True, random_state=seed)
import pandas as pd
cv1 = xgb.cv(params, dm, num_boost_round=10, nfold=nfolds, seed=seed) cv1 = xgb.cv(params, dm, num_boost_round=10, nfold=nfolds, seed=seed)
cv2 = xgb.cv(params, dm, num_boost_round=10, folds=skf, seed=seed) cv2 = xgb.cv(params, dm, num_boost_round=10, folds=skf, seed=seed)
cv3 = xgb.cv(params, dm, num_boost_round=10, nfold=nfolds, stratified=True, seed=seed) cv3 = xgb.cv(params, dm, num_boost_round=10, nfold=nfolds, stratified=True, seed=seed)
assert cv1.shape[0] == cv2.shape[0] and cv2.shape[0] == cv3.shape[0] assert cv1.shape[0] == cv2.shape[0] and cv2.shape[0] == cv3.shape[0]
assert cv2.iloc[-1,0] == cv3.iloc[-1,0] assert cv2.iloc[-1, 0] == cv3.iloc[-1, 0]

3
tests/travis/run_test.sh
View File

@ -52,6 +52,9 @@ if [ ${TASK} == "python_lightweight_test" ]; then
conda install numpy scipy nose conda install numpy scipy nose
python -m pip install graphviz python -m pip install graphviz
python -m nose tests/python/test_basic*.py || exit -1 python -m nose tests/python/test_basic*.py || exit -1
python -m pip install flake8
flake8 --ignore E501 python-package || exit -1
flake8 --ignore E501 tests/python || exit -1
exit 0 exit 0
fi fi