def setup_grid():
h2o.remove_all()
hyper_parameters = OrderedDict()
hyper_parameters["learn_rate"] = [0.1, 0.05, 0.01]
hyper_parameters["ntrees"] = [1, 3, 5]
gs = H2OGridSearch(H2OGradientBoostingEstimator, hyper_params=hyper_parameters)
return gs
def pca_pubdev_4167_OOM():
"""
This pyunit is written to make sure PCA works with customer data. It is mainly used by customer to verify
PCA operations and not to be used as a regular test since I do not want to expose customer data.
"""
h2o.remove_all()
transform_types = ["NONE", "STANDARDIZE", "NORMALIZE", "DEMEAN", "DESCALE"] # make sure we check all tranforms
transformN = transform_types[randint(0, len(transform_types)-1)]
print("transform used on dataset is {0}.\n".format(transformN))
training_data = h2o.import_file(path=pyunit_utils.locate("/Users/wendycwong/gitBackup/SDatasets/pubdev_4167_Avkash/m120K.tar")) # Nidhi: import may not work
gramSVDPCA = H2OPCA(k=training_data.ncols, transform=transformN)
gramSVDPCA.train(x=list(range(0, training_data.ncols)), training_frame=training_data)
powerSVDPCA = H2OPCA(k=training_data.ncols, transform=transformN, pca_method="Power")
powerSVDPCA.train(x=list(range(0, training_data.ncols)), training_frame=training_data)
# compare singular values and stuff between power and GramSVD methods
print("@@@@@@ Comparing eigenvalues between GramSVD and Power...\n")
pyunit_utils.assert_H2OTwoDimTable_equal(gramSVDPCA._model_json["output"]["importance"],
powerSVDPCA._model_json["output"]["importance"],
["Standard deviation", "Cumulative Proportion", "Cumulative Proportion"],
tolerance=1e-5, check_all=False)
print("@@@@@@ Comparing eigenvectors between GramSVD and Power...\n")
# compare singular vectors
pyunit_utils.assert_H2OTwoDimTable_equal(gramSVDPCA._model_json["output"]["eigenvectors"],
powerSVDPCA._model_json["output"]["eigenvectors"],
powerSVDPCA._model_json["output"]["names"], tolerance=1e-1,
check_sign=True)
def grid_resume():
train = h2o.import_file(path=pyunit_utils.locate("smalldata/iris/iris_wheader.csv"))
# Run GBM Grid Search
ntrees_opts = [1,5]
hyper_parameters = OrderedDict()
hyper_parameters["ntrees"] = ntrees_opts
print("GBM grid with the following hyper_parameters:", hyper_parameters)
export_dir = pyunit_utils.locate("results")
gs = H2OGridSearch(H2OGradientBoostingEstimator, hyper_params=hyper_parameters,
export_checkpoints_dir=export_dir)
gs.train(x=list(range(4)), y=4, training_frame=train)
grid_id = gs.grid_id
old_grid_model_count = len(gs.model_ids)
print("Baseline grid has %d models" % old_grid_model_count)
h2o.remove_all()
train = h2o.import_file(path=pyunit_utils.locate("smalldata/iris/iris_wheader.csv"))
grid = h2o.load_grid(export_dir + "/" + grid_id)
assert grid is not None
assert len(grid.model_ids) == old_grid_model_count
grid.train(x=list(range(4)), y=4, training_frame=train)
assert len(grid.model_ids) == old_grid_model_count
print("Newly grained grid has %d models" % len(grid.model_ids))
for model_id in grid.model_ids:
model = h2o.get_model(model_id)
assert model is not None
def deepwater_checkpoint():
if not H2ODeepWaterEstimator.available(): return
## build a model
#frame = h2o.import_file(pyunit_utils.locate("bigdata/laptop/deepwater/imagenet/cat_dog_mouse.csv"))
frame = h2o.import_file(pyunit_utils.locate("smalldata/prostate/prostate.csv"))
frame.drop(0)
frame[1] = frame[1].asfactor()
print(frame.head(5))
model = H2ODeepWaterEstimator(epochs=50, learning_rate=1e-5, stopping_rounds=0, score_duty_cycle=1, train_samples_per_iteration=-1, score_interval=0)
model.train(y=1, training_frame=frame)
## save the model
model_path = h2o.save_model(model)
## delete everything - simulate cluster shutdown and restart
h2o.remove_all()
## reimport the model and the frame
model = h2o.load_model(model_path)
#frame = h2o.import_file(pyunit_utils.locate("bigdata/laptop/deepwater/imagenet/cat_dog_mouse.csv"))
frame = h2o.import_file(pyunit_utils.locate("smalldata/prostate/prostate.csv"))
frame.drop(0)
frame[1] = frame[1].asfactor()
## delete the checkpoint file
os.remove(model_path)
## continue training
model2 = H2ODeepWaterEstimator(epochs=100, learning_rate=1e-5, stopping_rounds=0,score_duty_cycle=1, train_samples_per_iteration=-1, score_interval=0, checkpoint=model.model_id)
model2.train(y=1, training_frame=frame)
model2.show()
def test_gam_transformed_frame_serialization():
h2o_data = h2o.import_file(path=pyunit_utils.locate(
"smalldata/glm_test/multinomial_10_classes_10_cols_10000_Rows_train.csv"
))
h2o_data["C1"] = h2o_data["C1"].asfactor()
h2o_data["C2"] = h2o_data["C2"].asfactor()
myX = ["C1", "C2"]
myY = "C11"
h2o_data["C11"] = h2o_data["C11"].asfactor()
h2o_model = H2OGeneralizedAdditiveEstimator(family="multinomial",
gam_columns=["C6", "C7", "C8"],
keep_gam_cols=True,
scale=[1, 1, 1],
num_knots=[5, 5, 5])
h2o_model.train(x=myX, y=myY, training_frame=h2o_data)
gam_frame = h2o.get_frame(
h2o_model._model_json["output"]["gam_transformed_center_key"])
tmpdir = tempfile.mkdtemp()
filename = os.path.join(tmpdir, "gamXFrame.csv")
h2o.download_csv(gam_frame, filename)
model_path = h2o.save_model(h2o_model, tmpdir)
h2o.remove_all()
loaded_model = h2o.load_model(model_path)
gam_frame_loaded = h2o.get_frame(
loaded_model._model_json["output"]["gam_transformed_center_key"])
gam_frame_original = h2o.import_file(filename)
pyunit_utils.compare_frames_local(gam_frame_loaded[2:15],
gam_frame_original[2:15],
prob=1,
tol=1e-6)
print("Test completed.")
def glrm_mojo():
h2o.remove_all()
NTESTROWS = 200 # number of test dataset rows
df = pyunit_utils.random_dataset("regression") # generate random dataset
train = df[NTESTROWS:, :]
test = df[:NTESTROWS, :]
x = df.names
transform_types = ["NONE", "STANDARDIZE", "NORMALIZE", "DEMEAN", "DESCALE"]
transformN = transform_types[randint(0, len(transform_types)-1)]
# build a GLRM model with random dataset generated earlier
glrmModel = H2OGeneralizedLowRankEstimator(k=3, transform=transformN, max_iterations=10)
glrmModel.train(x=x, training_frame=train)
glrmTrainFactor = h2o.get_frame(glrmModel._model_json['output']['representation_name'])
assert glrmTrainFactor.nrows==train.nrows, \
"X factor row number {0} should equal training row number {1}.".format(glrmTrainFactor.nrows, train.nrows)
save_GLRM_mojo(glrmModel) # ave mojo model
MOJONAME = pyunit_utils.getMojoName(glrmModel._id)
TMPDIR = os.path.normpath(os.path.join(os.path.dirname(os.path.realpath('__file__')), "..", "results", MOJONAME))
h2o.download_csv(test[x], os.path.join(TMPDIR, 'in.csv')) # save test file, h2o predict/mojo use same file
pred_h2o, pred_mojo = pyunit_utils.mojo_predict(glrmModel, TMPDIR, MOJONAME, glrmReconstruct=True) # save mojo predict
for col in range(pred_h2o.ncols):
if pred_h2o[col].isfactor():
pred_h2o[col] = pred_h2o[col].asnumeric()
print("Comparing mojo predict and h2o predict...")
pyunit_utils.compare_frames_local(pred_h2o, pred_mojo, 1, tol=1e-10)
frameID, mojoXFactor = pyunit_utils.mojo_predict(glrmModel, TMPDIR, MOJONAME, glrmReconstruct=False) # save mojo XFactor
glrmTestFactor = h2o.get_frame("GLRMLoading_"+frameID) # store the x Factor for new test dataset
print("Comparing mojo x Factor and model x Factor ...")
pyunit_utils.compare_frames_local(glrmTestFactor, mojoXFactor, 1, tol=1e-10)
def grid_resume():
train = h2o.import_file(path=pyunit_utils.locate("smalldata/iris/iris_wheader.csv"))
# Run GBM Grid Search
ntrees_opts = [1, 3]
learn_rate_opts = [0.1, .05]
hyper_parameters = OrderedDict()
hyper_parameters["learn_rate"] = learn_rate_opts
hyper_parameters["ntrees"] = ntrees_opts
print("GBM grid with the following hyper_parameters:", hyper_parameters)
export_dir = pyunit_utils.locate("results")
gs = H2OGridSearch(H2OGradientBoostingEstimator, hyper_params=hyper_parameters)
gs.train(x=list(range(4)), y=4, training_frame=train)
grid_id = gs.grid_id
old_grid_model_count = len(gs.model_ids)
print("Baseline grid has %d models" % old_grid_model_count)
saved_path = h2o.save_grid(export_dir, grid_id)
h2o.remove_all();
train = h2o.import_file(path=pyunit_utils.locate("smalldata/iris/iris_wheader.csv"))
grid = h2o.load_grid(saved_path)
assert grid is not None
assert len(grid.model_ids) == old_grid_model_count
# Modify the hyperspace - should add new models to the grid
hyper_parameters["ntrees"] = [2,5]
grid = H2OGridSearch(H2OGradientBoostingEstimator, hyper_params=hyper_parameters, grid_id = grid.grid_id)
grid.train(x=list(range(4)), y=4, training_frame=train)
print("Newly grained grid has %d models" % len(grid.model_ids))
assert len(grid.model_ids) == 2 * old_grid_model_count
for model_id in grid.model_ids:
model = h2o.get_model(model_id)
assert model is not None
def deeplearning_mojo_pojo():
h2o.remove_all()
params = set_params() # set deeplearning model parameters
df = random_dataset(PROBLEM) # generate random dataset
train = df[NTESTROWS:, :]
test = df[:NTESTROWS, :]
x = list(set(df.names) - {"response"})
try:
deeplearningModel = build_save_model(params, x, train) # build and save mojo model
h2o.download_csv(test[x], os.path.join(TMPDIR, 'in.csv')) # save test file, h2o predict/mojo use same file
pred_h2o, pred_mojo = pyunit_utils.mojo_predict(deeplearningModel, TMPDIR, MOJONAME) # load model and perform predict
pred_pojo = pyunit_utils.pojo_predict(deeplearningModel, TMPDIR, MOJONAME)
h2o.save_model(deeplearningModel, path=TMPDIR, force=True) # save model for debugging
print("Comparing mojo predict and h2o predict...")
pyunit_utils.compare_numeric_frames(pred_h2o, pred_mojo, 0.1, tol=1e-10) # make sure operation sequence is preserved from Tomk
print("Comparing pojo predict and h2o predict...")
pyunit_utils.compare_numeric_frames(pred_mojo, pred_pojo, 0.1, tol=1e-10)
except Exception as ex:
print("*************** ERROR and type is ")
print(str(type(ex)))
print(ex)
if "AssertionError" in str(type(ex)): # only care if there is an AssertionError, ignore the others
sys.exit(1)
def glrm_mojo():
h2o.remove_all()
train = h2o.import_file(pyunit_utils.locate("smalldata/glrm_test/pubdev_5858_glrm_mojo_train.csv"))
test = h2o.import_file(pyunit_utils.locate("smalldata/glrm_test/pubdev_5858_glrm_mojo_test.csv"))
predict_10iter = h2o.import_file(pyunit_utils.locate("smalldata/glrm_test/pubdev_5858_glrm_predict_10iter.csv"))
predict_1iter = h2o.import_file(pyunit_utils.locate("smalldata/glrm_test/pubdev_5858_glrm_predict_1iter.csv"))
x = train.names
transformN = "STANDARDIZE"
# build a GLRM model with random dataset generated earlier
glrmModel = H2OGeneralizedLowRankEstimator(k=3, transform=transformN, max_iterations=10, seed=1234, init="random")
glrmModel.train(x=x, training_frame=train)
glrmTrainFactor = h2o.get_frame(glrmModel._model_json['output']['representation_name'])
assert glrmTrainFactor.nrows==train.nrows, \
"X factor row number {0} should equal training row number {1}.".format(glrmTrainFactor.nrows, train.nrows)
save_GLRM_mojo(glrmModel) # save mojo model
MOJONAME = pyunit_utils.getMojoName(glrmModel._id)
TMPDIR = os.path.normpath(os.path.join(os.path.dirname(os.path.realpath('__file__')), "..", "results", MOJONAME))
h2o.download_csv(test[x], os.path.join(TMPDIR, 'in.csv')) # save test file, h2o predict/mojo use same file
# test and make sure setting the iteration number did not screw up the prediction
predID, pred_mojo = pyunit_utils.mojo_predict(glrmModel, TMPDIR, MOJONAME, glrmIterNumber=100) # save mojo predict
pred_h2o = h2o.get_frame("GLRMLoading_"+predID)
print("Comparing mojo x Factor and model x Factor for 100 iterations")
pyunit_utils.compare_frames_local(pred_h2o, pred_mojo, 1, tol=1e-10)
predID, pred_mojo = pyunit_utils.mojo_predict(glrmModel, TMPDIR, MOJONAME, glrmIterNumber=1) # save mojo predict
print("Comparing mojo x Factor and model x Factor for 1 iterations")
pyunit_utils.compare_frames_local(predict_1iter, pred_mojo, 1, tol=1e-10)
predID, pred_mojo = pyunit_utils.mojo_predict(glrmModel, TMPDIR, MOJONAME, glrmIterNumber=10) # save mojo predict
print("Comparing mojo x Factor and model x Factor for 10 iterations")
pyunit_utils.compare_frames_local(predict_10iter, pred_mojo, 1, tol=1e-10)
def test_modelselection_backward_serialization():
d = h2o.import_file(path=pyunit_utils.locate("smalldata/logreg/prostate.csv"))
y = "GLEASON"
x = ["ID","AGE","RACE","CAPSULE","DCAPS","PSA","VOL","DPROS"]
# make sure duplicate runs produce same results
model_backward = modelSelection(seed=12345, mode="backward", family='negativebinomial', link="log",alpha=0.5,
lambda_=0, theta=0.01)
model_backward.train(training_frame=d, x=x, y=y)
model_backward2 = modelSelection(seed=12345, mode="backward", family='negativebinomial', link="log",alpha=0.5,
lambda_=0, theta=0.01)
model_backward2.train(training_frame=d, x=x, y=y)
result = model_backward.result() # get result frame
result2 = model_backward.result() # get result frame
pyunit_utils.compare_frames_local(result[2:5], result2[2:5], prob=1.0) # compare result from both models and they should the same
num_models = result.nrows # number of models built
one_model = h2o.get_model(result["model_id"][num_models-1, 0])
predict_frame = one_model.predict(d)
tmpdir = tempfile.mkdtemp()
file_dir = os.path.join(tmpdir, "predict.csv")
h2o.download_csv(predict_frame, file_dir) # save one scoring frame
model_path_backward = model_backward.download_model(tmpdir) # store the model
h2o.remove_all()
d = h2o.import_file(path=pyunit_utils.locate("smalldata/logreg/prostate.csv"))
loaded_backward_model = h2o.load_model(model_path_backward)
result_frame_backward = loaded_backward_model.result()
model_from_frame_backward = h2o.get_model(result_frame_backward["model_id"][num_models-1, 0])
pred_frame_backward = model_from_frame_backward.predict(d)
pred_frame_model = h2o.import_file(file_dir)
pyunit_utils.compare_frames_local(pred_frame_backward, pred_frame_model, prob=1.0)
def check_big_merge():
h2o.remove_all()
nrow = 1000000
ncol = 2
iRange = 100000
frame1 = h2o.create_frame(rows=nrow,
cols=ncol,
integer_fraction=1,
seed=12345,
integer_range=iRange,
missing_fraction=0.0)
frame2 = h2o.create_frame(rows=nrow,
cols=ncol,
integer_fraction=1,
seed=54321,
integer_range=iRange,
missing_fraction=0.0)
frame1.set_names(["C1", "C2"])
frame2.set_names(["C1", "C3"])
mergedExact = frame1.merge(frame2,
by_x=["C1"],
by_y=["C1"],
all_x=False,
all_y=False)
mergedLeft = frame1.merge(frame2, by_x=["C1"], by_y=["C1"], all_x=True)
assert mergedExact.nrow < mergedLeft.nrow, "Expected row numbers are wrong"
def glm_binomial_mojo_pojo():
h2o.remove_all()
NTESTROWS = 200 # number of test dataset rows
PROBLEM = "binomial"
params = set_params() # set deeplearning model parameters
df = pyunit_utils.random_dataset(PROBLEM) # generate random dataset
train = df[NTESTROWS:, :]
test = df[:NTESTROWS, :]
x = list(set(df.names) - {"response"})
TMPDIR = tempfile.mkdtemp()
glmBinomialModel = pyunit_utils.build_save_model_generic(
params, x, train, "response", "glm",
TMPDIR) # build and save mojo model
MOJONAME = pyunit_utils.getMojoName(glmBinomialModel._id)
h2o.download_csv(test[x], os.path.join(
TMPDIR, 'in.csv')) # save test file, h2o predict/mojo use same file
pred_h2o, pred_mojo = pyunit_utils.mojo_predict(
glmBinomialModel, TMPDIR, MOJONAME) # load model and perform predict
h2o.download_csv(pred_h2o, os.path.join(TMPDIR, "h2oPred.csv"))
pred_pojo = pyunit_utils.pojo_predict(glmBinomialModel, TMPDIR, MOJONAME)
print("Comparing mojo predict and h2o predict...")
pyunit_utils.compare_frames_local(
pred_h2o, pred_mojo, 0.1, tol=1e-10
) # make sure operation sequence is preserved from Tomk h2o.save_model(glmOrdinalModel, path=TMPDIR, force=True) # save model for debugging
print("Comparing pojo predict and h2o predict...")
pyunit_utils.compare_frames_local(pred_mojo, pred_pojo, 0.1, tol=1e-10)
def test_anovaglm_serialization():
train = h2o.import_file(path=pyunit_utils.locate(
"smalldata/prostate/prostate_complete.csv.zip"))
y = 'CAPSULE'
x = ['AGE', 'VOL', 'DCAPS']
train[y] = train[y].asfactor()
anovaglm_model = anovaglm(family='binomial',
lambda_=0,
missing_values_handling="skip")
anovaglm_model.train(x=x, y=y, training_frame=train)
tmpdir = tempfile.mkdtemp()
model_path = anovaglm_model.download_model(tmpdir)
result_frame_filename = os.path.join(tmpdir, "result_frame.csv")
h2o.download_csv(anovaglm_model.result(), result_frame_filename)
h2o.remove_all()
result_frame_original = h2o.import_file(result_frame_filename)
loaded_anovaglm_model = h2o.load_model(model_path)
result_frame_loaded = loaded_anovaglm_model.result()
for cind in list(range(0, result_frame_original.ncols)):
for rind in list(range(0, result_frame_original.nrows)):
if result_frame_original.type(cind) == 'real':
assert abs(result_frame_original[rind, cind]-result_frame_loaded[rind, cind]) < 1e-6, \
"Expected: {0}. Actual: {1}".format(result_frame_original[rind, cind], result_frame_loaded[rind, cind])
else:
assert result_frame_original[rind, cind]==result_frame_loaded[rind, cind], \
"Expected: {0}. Actual: {1}".format(result_frame_original[rind, cind], result_frame_loaded[rind, cind])
def grid_export_with_cv():
train = h2o.import_file(path=pyunit_utils.locate("smalldata/iris/iris_wheader.csv"))
# Run GBM Grid Search
hyper_parameters = OrderedDict()
hyper_parameters["ntrees"] = [1, 2]
# train with CV
gs = H2OGridSearch(H2OGradientBoostingEstimator(nfolds=2, keep_cross_validation_predictions=True, seed=42),
hyper_params=hyper_parameters)
gs.train(x=list(range(4)), y=4, training_frame=train)
holdout_frame_ids = map(lambda m: m.cross_validation_holdout_predictions().frame_id, gs.models)
export_dir = pyunit_utils.locate("results")
saved_path = h2o.save_grid(export_dir, gs.grid_id, export_cross_validation_predictions=True)
h2o.remove_all()
grid = h2o.load_grid(saved_path)
assert grid is not None
for holdout_frame_id in holdout_frame_ids:
assert h2o.get_frame(holdout_frame_id) is not None
train = h2o.import_file(path=pyunit_utils.locate("smalldata/iris/iris_wheader.csv"))
stack = H2OStackedEnsembleEstimator(base_models=grid.model_ids)
stack.train(x=list(range(4)), y=4, training_frame=train)
predicted = stack.predict(train)
assert predicted.nrow == train.nrow
def gam_gaussian_mojo():
h2o.remove_all()
NTESTROWS = 200 # number of test dataset rows
PROBLEM="gaussian"
params = set_params() # set deeplearning model parameters
df = pyunit_utils.random_dataset(PROBLEM, missing_fraction=0.001) # generate random dataset
dfnames = df.names
# add GAM specific parameters
params["gam_columns"] = []
params["scale"] = []
count = 0
num_gam_cols = 3 # maximum number of gam columns
for cname in dfnames:
if not(cname == 'response') and (str(df.type(cname)) == "real"):
params["gam_columns"].append(cname)
params["scale"].append(0.001)
count = count+1
if (count >= num_gam_cols):
break
train = df[NTESTROWS:, :]
test = df[:NTESTROWS, :]
x = list(set(df.names) - {"response"})
TMPDIR = tempfile.mkdtemp()
gamGaussianModel = pyunit_utils.build_save_model_generic(params, x, train, "response", "gam", TMPDIR) # build and save mojo model
MOJONAME = pyunit_utils.getMojoName(gamGaussianModel._id)
h2o.download_csv(test[x], os.path.join(TMPDIR, 'in.csv')) # save test file, h2o predict/mojo use same file
pred_h2o, pred_mojo = pyunit_utils.mojo_predict(gamGaussianModel, TMPDIR, MOJONAME) # load model and perform predict
h2o.download_csv(pred_h2o, os.path.join(TMPDIR, "h2oPred.csv"))
print("Comparing mojo predict and h2o predict...")
pyunit_utils.compare_frames_local(pred_h2o, pred_mojo, 0.1, tol=1e-10) # make sure operation sequence is preserved from Tomk h2o.save_model(glmOrdinalModel, path=TMPDIR, force=True) # save model for debugging
def test_frame_reload():
work_dir = tempfile.mkdtemp()
iris = h2o.import_file(
path=pyunit_utils.locate("smalldata/iris/iris_wheader.csv"))
df_key = iris.key
df_pd_orig = iris.as_data_frame()
iris.save(work_dir)
try:
iris.save(work_dir, force=False) # fails because file exists
except H2OResponseError as e:
assert e.args[0].exception_msg.startswith("File already exists")
try:
h2o.load_frame(df_key, work_dir,
force=False) # fails because frame exists
except H2OResponseError as e:
assert e.args[
0].exception_msg == "Frame Key<Frame> iris_wheader.hex already exists."
df_loaded_force = h2o.load_frame(df_key, work_dir)
h2o.remove(iris)
df_loaded = h2o.load_frame(df_key, work_dir, force=False)
df_pd_loaded_force = df_loaded_force.as_data_frame()
df_pd_loaded = df_loaded.as_data_frame()
assert df_pd_orig.equals(df_pd_loaded_force)
assert df_pd_orig.equals(df_pd_loaded)
# try running grid search on the frame
h2o.remove_all()
df_loaded = h2o.load_frame(df_key, work_dir)
hyper_parameters = OrderedDict()
hyper_parameters["ntrees"] = [5, 10, 20, 30]
grid_small = H2OGridSearch(H2OGradientBoostingEstimator,
hyper_params=hyper_parameters)
grid_small.train(x=list(range(4)), y=4, training_frame=df_loaded)
assert len(grid_small.models) == 4
def check_story(story_name, paragraphs):
h2o.remove_all()
h2o.log_and_echo("------------------------------------------------------------")
h2o.log_and_echo("")
h2o.log_and_echo("CHECKING: {0}".format(story_name))
h2o.log_and_echo("")
h2o.log_and_echo("------------------------------------------------------------")
# 1. Combine the related, individual code paragraphs into a single, coherent python story
story = []
for p in paragraphs:
with open(p, "r") as f: story = story + f.readlines()
# 2. Execute the story
# first, remove any h2o.init calls
remove_lines = []
for idx, l in enumerate(story):
if "h2o.init" in l: remove_lines.append(idx)
story = [i for j, i in enumerate(story) if j not in remove_lines]
# write the story that will be executed to the results directory for future reference
story_file = os.path.join(results_dir(), test_name()+"."+story_name+".code")
with open(story_file, 'w') as f: f.writelines(story)
# run it
with open(story_file, "r") as s: booklet = s.read()
booklet_c = compile(booklet, '<string>', 'exec')
p = {}
exec(booklet_c, p)
def save_load_mode_with_cv():
prostate = h2o.import_file(
pyunit_utils.locate("smalldata/prostate/prostate.csv"))
prostate["CAPSULE"] = prostate["CAPSULE"].asfactor()
prostate_gbm = H2OGradientBoostingEstimator(
nfolds=2, keep_cross_validation_predictions=True)
prostate_gbm.train(x=["AGE", "RACE", "PSA", "DCAPS"],
y="CAPSULE",
training_frame=prostate)
path = pyunit_utils.locate("results")
model_path = h2o.save_model(prostate_gbm,
path=path,
force=True,
export_cross_validation_predictions=True)
assert os.path.isfile(
model_path
), "Expected model artifact {0} to exist, but it does not.".format(
model_path)
h2o.remove_all()
prostate_gbm_reloaded = h2o.load_model(model_path)
assert isinstance(prostate_gbm_reloaded, H2OGradientBoostingEstimator), \
"Expected H2OGradientBoostingEstimator, but got {0}".format(prostate_gbm_reloaded)
holdout_frame_id = prostate_gbm.cross_validation_holdout_predictions(
).frame_id
assert h2o.get_frame(holdout_frame_id) is not None
def deeplearning_mojo_pojo():
h2o.remove_all()
params = set_params() # set deeplearning model parameters
df = random_dataset(PROBLEM) # generate random dataset
train = df[NTESTROWS:, :]
test = df[:NTESTROWS, :]
x = list(set(df.names) - {"response"})
try:
deeplearningModel = build_save_model(
params, x, train) # build and save mojo model
h2o.download_csv(test[x], os.path.join(
TMPDIR,
'in.csv')) # save test file, h2o predict/mojo use same file
pred_h2o, pred_mojo = pyunit_utils.mojo_predict(
deeplearningModel, TMPDIR,
MOJONAME) # load model and perform predict
# pred_pojo = pyunit_utils.pojo_predict(deeplearningModel, TMPDIR, MOJONAME)
h2o.save_model(deeplearningModel, path=TMPDIR,
force=True) # save model for debugging
print("Comparing mojo predict and h2o predict...")
pyunit_utils.compare_numeric_frames(pred_h2o, pred_mojo, 0.1, tol=1e-6)
# print("Comparing pojo predict and h2o predict...")
# pyunit_utils.compare_numeric_frames(pred_mojo, pred_pojo, 0.1, tol=1e-6)
except Exception as ex:
print("*************** ERROR and type is ")
print(str(type(ex)))
print(ex)
if "AssertionError" in str(
type(ex)
): # only care if there is an AssertionError, ignore the others
sys.exit(1)
def check_story(story_name, paragraphs):
h2o.remove_all()
h2o.log_and_echo("------------------------------------------------------------")
h2o.log_and_echo("")
h2o.log_and_echo("CHECKING: {0}".format(story_name))
h2o.log_and_echo("")
h2o.log_and_echo("------------------------------------------------------------")
# 1. Combine the related, individual code paragraphs into a single, coherent python story
story = []
for p in paragraphs:
with open(p, "r") as f: story = story + f.readlines()
# 2. Execute the story
# first, remove any h2o.init calls
remove_lines = []
for idx, l in enumerate(story):
if "h2o.init" in l: remove_lines.append(idx)
story = [i for j, i in enumerate(story) if j not in remove_lines]
# write the story that will be executed to the results directory for future reference
story_file = os.path.join(results_dir(), test_name()+"."+story_name+".code")
with open(story_file, 'w') as f: f.writelines(story)
# run it
with open(story_file, "r") as s: booklet = s.read()
booklet_c = compile(booklet, '<string>', 'exec')
p = {}
exec(booklet_c, p)
def train_with_h2o_full_data(time):
import h2o
from h2o.automl import H2OAutoML
h2o.init()
h2o.remove_all()
raw_train_df = h2o.import_file(
path='data/train_dataset/train_dataset_temp.csv')
raw_test_df = h2o.import_file(
path='data/test_dataset/test_dataset_temp.csv')
train_df = raw_train_df[:, 1:]
y = 'var_29'
col_list = train_df.columns
print(col_list)
x = col_list[:-1]
print('x:', x)
print('y:', y)
splits = train_df.split_frame(ratios=[0.9], seed=1)
train = splits[0]
test = splits[1]
#part 50000行
# aml1 = H2OAutoML(max_runtime_secs=time,balance_classes=False,stopping_tolerance=0.005,stopping_rounds=50,sort_metric='MAE',stopping_metric='MAE',seed=2019, project_name="part_data_train")
# aml1.train(x=x,y=y,training_frame=train,leaderboard_frame=test)
#full data train
aml2 = H2OAutoML(max_runtime_secs=time,
balance_classes=False,
stopping_tolerance=0.005,
stopping_rounds=50,
sort_metric='MAE',
stopping_metric='MAE',
seed=2019,
project_name="full_data_train")
aml2.train(x=x, y=y, training_frame=train_df)
# path1=h2o.save_model(model=aml1,path='models',force=True)
# path2=h2o.save_model(model=aml2,path='models',force=True)
# print(aml1.leaderboard)
print('++++++++++++++++++++++')
print(aml2.leaderboard)
# ans1=aml1.predict(raw_test_df[:,1:])
ans2 = aml2.predict(raw_test_df[:, 1:])
# print(ans1)
print(ans2)
# ans1=ans1.as_data_frame()
ans2 = ans2.as_data_frame()
# ans1.to_csv('data/ans1.csv',index=False)
ans2.to_csv('data/ans2_time_{}.csv'.format(str(time)), index=False)
# res1=pd.DataFrame()
temp = pd.read_csv('data/test_dataset/test_dataset_temp.csv')
# res1['id']=temp['var_0']
# res1['score']=ans1.values
# res1.to_csv('data/h2o_pred_submission_v1.csv',index=False)
res2 = pd.DataFrame()
res2['id'] = temp['var_0']
res2['score'] = ans2.values
res2['score'] = res2['score'].apply(lambda x: int(x)
if (x - int(x)) < 0.5 else int(x) + 1)
res2.to_csv('data/h2o_pred_submission_int_v2_time_{}.csv'.format(time),
index=False)
def setup_grid():
h2o.remove_all()
hyper_parameters = OrderedDict()
hyper_parameters["learn_rate"] = [0.1, 0.05, 0.01]
hyper_parameters["ntrees"] = [1, 3, 5]
gs = H2OGridSearch(H2OGradientBoostingEstimator,
hyper_params=hyper_parameters)
return gs
def impute_missing_values(data_frame, columns):
if not isinstance(data_frame, pd.DataFrame):
return
if not isinstance(columns, list):
return
# Impute summary
impute_summary = {}
# result frame
result_frame = pd.DataFrame(data_frame)
# Start h2o server
h2o.init(max_mem_size_GB=5)
for column in columns:
print "Processing :", column
# Defining columns
response_column = column
training_columns = list(data_frame.columns)
training_columns.remove(response_column)
# Creating h2o frame
training_frame = h2o.H2OFrame(data_frame)
training_frame.set_names(list(data_frame.columns))
# Defining model
model = H2ORandomForestEstimator(ntrees=75,
max_depth=25,
nbins=25,
binomial_double_trees=True,
nfolds=10)
model.train(x=training_columns,
y=response_column,
training_frame=training_frame)
# Predict values
predictions = model.predict(test_data=training_frame)
predictions = list(map(float, h2OColumnToList(predictions)))
# Add predictions to the result frame
result_frame[column] = predictions
actual = data_frame[column]
predicted = result_frame[column]
rmse = sqrt(mean_squared_error(actual, predicted))
impute_summary[column] = ('RMSE', rmse)
# Removing all processes
h2o.remove_all()
# Displaying impute summary
for key in impute_summary:
print impute_summary[key], key
return result_frame
def test_stacked_ensemble_is_able_to_use_imported_base_models():
import tempfile, shutil, glob
train = h2o.import_file(pu.locate("smalldata/iris/iris_train.csv"))
test = h2o.import_file(pu.locate("smalldata/iris/iris_test.csv"))
x = train.columns
y = "species"
x.remove(y)
nfolds = 2
gbm = H2OGradientBoostingEstimator(nfolds=nfolds,
fold_assignment="Modulo",
keep_cross_validation_predictions=True)
gbm.train(x=x, y=y, training_frame=train)
drf = H2ORandomForestEstimator(nfolds=nfolds,
fold_assignment="Modulo",
keep_cross_validation_predictions=True)
drf.train(x=x, y=y, training_frame=train)
se = H2OStackedEnsembleEstimator(training_frame=train,
validation_frame=test,
base_models=[gbm.model_id, drf.model_id])
se.train(x=x, y=y, training_frame=train)
assert len(se.base_models) == 2
TMP_DIR = tempfile.mkdtemp()
try:
h2o.save_model(gbm, TMP_DIR + "/gbm.model")
h2o.save_model(drf, TMP_DIR + "/drf.model")
gbm_holdout_id = gbm.cross_validation_holdout_predictions().frame_id
drf_holdout_id = drf.cross_validation_holdout_predictions().frame_id
h2o.export_file(gbm.cross_validation_holdout_predictions(), TMP_DIR + "/gbm.holdout")
h2o.export_file(drf.cross_validation_holdout_predictions(), TMP_DIR + "/drf.holdout")
h2o.remove_all()
h2o.import_file(TMP_DIR + "/gbm.holdout", gbm_holdout_id)
h2o.import_file(TMP_DIR + "/drf.holdout", drf_holdout_id)
gbm = h2o.upload_model(glob.glob(TMP_DIR + "/gbm.model/*")[0])
drf = h2o.upload_model(glob.glob(TMP_DIR + "/drf.model/*")[0])
train = h2o.import_file(pu.locate("smalldata/iris/iris_train.csv"), "some_other_name_of_training_frame")
test = h2o.import_file(pu.locate("smalldata/iris/iris_test.csv"), "some_other_name_of_test_frame")
x = train.columns
y = "species"
x.remove(y)
se_loaded = H2OStackedEnsembleEstimator(training_frame=train,
validation_frame=test,
base_models=[gbm.model_id, drf.model_id])
se_loaded.train(x=x, y=y, training_frame=train)
assert len(se_loaded.base_models) == 2
finally:
shutil.rmtree(TMP_DIR)
def validacion_r(modelo, hyper_parameters, datos, variables, semilla=1234):
h2o.init(max_mem_size=14)
train = h2o.H2OFrame(datos[0])
tipificar_h2o(train)
splits = train.split_frame(ratios=[0.7], seed=semilla)
gs = H2OGridSearch(modelo, hyper_params=hyper_parameters)
gs.train(x=variables, y="Tendencia", training_frame=splits[0])
resultados=procesamiento_resultados_binario(gs,splits,datos)
h2o.remove_all()
return(resultados)
def deeplearning_mojo_pojo():
h2o.remove_all()
problemtypes = ["regression", "binomial", "multinomial"]
autoEncoderOn = [True, False]
for encoderOn in autoEncoderOn:
for problem in problemtypes:
print("AutoEncoderOn is: {0} and problem type is: {1}".format(encoderOn, problem))
random.seed(9876) # set python random seed
runComparisonTests(encoderOn, problem)
def h2o_test_setup(sys_args):
h2o_py_dir = os.path.realpath(
os.path.join(os.path.dirname(os.path.realpath(__file__)), ".."))
h2o_docs_dir = os.path.realpath(
os.path.join(os.path.dirname(os.path.realpath(__file__)), "..", "..",
"h2o-docs"))
parse_args(sys_args)
sys.path.insert(1, h2o_py_dir)
import h2o
from tests import pyunit_utils, pydemo_utils, pybooklet_utils
for pkg in (pyunit_utils, pybooklet_utils):
setattr(pkg, '__on_hadoop__', _ON_HADOOP_)
setattr(pkg, '__hadoop_namenode__', _HADOOP_NAMENODE_)
setattr(pkg, '__test_name__', _TEST_NAME_)
setattr(pkg, '__results_dir__', _RESULTS_DIR_)
if _IS_PYUNIT_ or _IS_IPYNB_ or _IS_PYBOOKLET_ or _IS_PYDEMO_:
pass
else:
raise (
EnvironmentError,
"Unrecognized test type. Must be of type ipynb, pydemo, pyunit, or pybooklet, but got: "
"{0}".format(_TEST_NAME_))
print("[{0}] {1}\n".format(
strftime("%Y-%m-%d %H:%M:%S", gmtime()),
"Connect to h2o on IP: {0} PORT: {1}".format(_H2O_IP_, _H2O_PORT_)))
auth = None
if _LDAP_USER_NAME_ is not None and _LDAP_PASSWORD_ is not None:
auth = (_LDAP_USER_NAME_, _LDAP_PASSWORD_)
h2o.connect(ip=_H2O_IP_, port=_H2O_PORT_, verbose=False, auth=auth)
h2o.utils.config.H2OConfigReader.get_config(
)["general.allow_breaking_changes"] = True
#rest_log = os.path.join(_RESULTS_DIR_, "rest.log")
#h2o.start_logging(rest_log)
#print "[{0}] {1}\n".format(strftime("%Y-%m-%d %H:%M:%S", gmtime()), "Started rest logging in: {0}".format(rest_log))
h2o.log_and_echo(
"------------------------------------------------------------")
h2o.log_and_echo("")
h2o.log_and_echo("STARTING TEST: " + _TEST_NAME_)
h2o.log_and_echo("")
h2o.log_and_echo(
"------------------------------------------------------------")
h2o.remove_all()
if _IS_IPYNB_: pydemo_utils.ipy_notebook_exec(_TEST_NAME_)
elif _IS_PYUNIT_: pyunit_utils.pyunit_exec(_TEST_NAME_)
elif _IS_PYBOOKLET_: pybooklet_utils.pybooklet_exec(_TEST_NAME_)
elif _IS_PYDEMO_: pydemo_utils.pydemo_exec(_TEST_NAME_)
def standalone_test(test):
h2o.init(strict_version_check=False)
h2o.remove_all()
h2o.log_and_echo("------------------------------------------------------------")
h2o.log_and_echo("")
h2o.log_and_echo("STARTING TEST")
h2o.log_and_echo("")
h2o.log_and_echo("------------------------------------------------------------")
test()
def standalone_test(test):
h2o.init(strict_version_check=False)
h2o.remove_all()
h2o.log_and_echo("------------------------------------------------------------")
h2o.log_and_echo("")
h2o.log_and_echo("STARTING TEST")
h2o.log_and_echo("")
h2o.log_and_echo("------------------------------------------------------------")
test()
def sort():
h2o.remove_all()
df = h2o.import_file(
pyunit_utils.locate(
"bigdata/laptop/jira/PUBDEV_6829_srot_bug_bigKey_part.csv.zip"))
t1 = time.time()
df1 = df.sort([1])
assert df1[0, 1] <= df1[1, 1], "Test failed: Sort bug."
print("Time taken to perform sort is {0}".format(time.time() - t1))
pyunit_utils.check_sorted_1_column(df1, 1, prob=0.00001,
ascending=True) # check some rows
def deepwater_lenet():
if not H2ODeepWaterEstimator.available(): return
frame = h2o.import_file(pyunit_utils.locate("bigdata/laptop/deepwater/imagenet/cat_dog_mouse.csv"))
print(frame.head(5))
model = H2ODeepWaterEstimator(epochs=100, learning_rate=1e-3, network='lenet', score_interval=0, train_samples_per_iteration=1000)
model.train(x=[0],y=1, training_frame=frame)
model.show()
error = model.model_performance(train=True).mean_per_class_error()
h2o.remove_all()
assert error < 0.1, "mean classification error is too high : " + str(error)
def test_rbind_summary():
h2o.remove_all()
df = h2o.H2OFrame([1, 2, 5.5], destination_frame="df") # original frame
dfr = h2o.H2OFrame([5.5, 1, 2],
destination_frame="dfr") # reversed row content
df1 = df[2, :]
df2 = df[:2, :]
summary = df1.summary(return_data=True)
df3 = df1.rbind(df2) # fixed
df3r = df2.rbind(df1)
compareFramesLocal(dfr, df3) # should contain 5.5, 1, 2
compareFramesLocal(df, df3r) # should contain 1,2,5.5
df1 = df[
3, :] # this will result in an NA since we do not have 4 rows in df.
dfr[0, 0] = float('nan')
df4 = df1.rbind(df2)
compareFramesLocal(df4, dfr) # should contain NA, 1, 2
# performing the same test with an additionl categorical column per Michalk request.
h2o.remove_all()
df = h2o.H2OFrame([[1, "a"], [2, "b"], [5.5, "c"]],
destination_frame="dfc") # original frame
df[1] = df[1].asfactor()
dfr = h2o.H2OFrame([[5.5, "c"], [1, "a"], [2, "b"]],
destination_frame="dfrc") # reversed row content
dfr[1] = df[1].asfactor(
) # this somehow switch the row content of the factor column to be alphabetical
dfr[0, 1] = 'c'
dfr[1, 1] = 'a'
dfr[2, 1] = 'b'
df1 = df[2, :]
df2 = df[:2, :]
summary = df1.summary(return_data=True)
df3 = df1.rbind(df2) # fixed
df3r = df2.rbind(df1)
compareFramesLocal(dfr, df3) # should contain 5.5, 1, 2
compareFramesLocal(df, df3r) # should contain 1,2,5.5
# copying test from Michalk
df1 = h2o.H2OFrame([[1, "a"], [2, "b"]])
df1[1] = df1[1].asfactor()
df2 = h2o.H2OFrame([[2.2, "b"], [1.1, "a"]])
df2[1] = df2[1].asfactor()
print(df1.summary())
print(df2.summary())
df3 = df1.rbind(df2)
assert df3.nrow==(df1.nrow+df2.nrow), "Expected rbind rows: {0}, actual rows: " \
"{1}".format(df1.nrow+df2.nrow, df3.nrow)
def h2o_test_setup(sys_args):
h2o_py_dir = os.path.realpath(
os.path.join(os.path.dirname(os.path.realpath(__file__)), ".."))
h2o_docs_dir = os.path.realpath(
os.path.join(os.path.dirname(os.path.realpath(__file__)), "..", "..",
"h2o-docs"))
parse_args(sys_args)
sys.path.insert(1, h2o_py_dir)
import h2o
from tests import pyunit_utils, pydemo_utils, pybooklet_utils
set_pyunit_pkg_attrs(pyunit_utils)
set_pybooklet_pkg_attrs(pybooklet_utils)
if _IS_PYUNIT_ or _IS_IPYNB_ or _IS_PYBOOKLET_ or _IS_PYDEMO_:
pass
else:
raise (
EnvironmentError,
"Unrecognized test type. Must be of type ipynb, pydemo, pyunit, or pybooklet, but got: "
"{0}".format(_TEST_NAME_))
print("[{0}] {1}\n".format(
strftime("%Y-%m-%d %H:%M:%S", gmtime()),
"Connect to h2o on IP: {0} PORT: {1}".format(_H2O_IP_, _H2O_PORT_)))
h2o.init(ip=_H2O_IP_,
port=_H2O_PORT_,
strict_version_check=False,
force_connect=_FORCE_CONNECT_)
h2o.utils.config.H2OConfigReader.get_config(
)["general.allow_breaking_changes"] = True
#rest_log = os.path.join(_RESULTS_DIR_, "rest.log")
#h2o.start_logging(rest_log)
#print "[{0}] {1}\n".format(strftime("%Y-%m-%d %H:%M:%S", gmtime()), "Started rest logging in: {0}".format(rest_log))
h2o.log_and_echo(
"------------------------------------------------------------")
h2o.log_and_echo("")
h2o.log_and_echo("STARTING TEST: " + _TEST_NAME_)
h2o.log_and_echo("")
h2o.log_and_echo(
"------------------------------------------------------------")
h2o.remove_all()
if _IS_IPYNB_: pydemo_utils.ipy_notebook_exec(_TEST_NAME_)
elif _IS_PYUNIT_: pyunit_utils.pyunit_exec(_TEST_NAME_)
elif _IS_PYBOOKLET_: pybooklet_utils.pybooklet_exec(_TEST_NAME_)
elif _IS_PYDEMO_: pydemo_utils.pydemo_exec(_TEST_NAME_)
def javamunge_assembly():
h2o.remove_all()
train = pyunit_utils.locate("bigdata/laptop/lending-club/LoanStats3a.csv")
test = pyunit_utils.locate("bigdata/laptop/lending-club/LoanStats3b.csv")
# lending-club munging assembly
print("Import and Parse data")
# Add "earliest_cr_line" and "issue_d" and cast as strings to aide Cliff's PR on 7/13
types = {"int_rate": "string", "revol_util": "string", "emp_length": "string", "earliest_cr_line": "string",
"issue_d": "string", "last_credit_pull_d": "factor"}
data = h2o.import_file(path=train, col_types=types)
test = h2o.import_file(path=test, col_types=data.types) ## use the same data types as the training set for the test set
test = test[[1,5,19,23,45,66,99,590,8903,9999,10001,23892,23893,50123],:]
test = h2o.assign(test,"test")
assembly = H2OAssembly(
steps=[
# munge int_rate column in place
# strip %, trim ws, convert to double
("intrate_rm_junk_char", H2OColOp(op=H2OFrame.gsub, col="int_rate", inplace=True, pattern="%", replacement="")), # strip %
("intrate_trim_ws", H2OColOp(op=H2OFrame.trim, col="int_rate", inplace=True)), # trim ws
("intrate_as_numeric", H2OColOp(op=H2OFrame.asnumeric, col="int_rate", inplace=True)), # string -> double
# munge the revol_util in the same way as the int_rate column
("revol_rm_junk_char", H2OColOp(op=H2OFrame.gsub, col="revol_util", inplace=True, pattern="%", replacement="")), # strip %
("revol_trim_ws", H2OColOp(op=H2OFrame.trim, col="revol_util", inplace=True)), # trim ws
("revol_as_numeric", H2OColOp(op=H2OFrame.asnumeric, col="revol_util", inplace=True)), # string -> double
# munge earliest_cr_line column (mm-YYYY format)
# split into Month and Year columns
("earliest_cr_line_split", H2OColOp(H2OFrame.strsplit, col="earliest_cr_line", inplace=False, new_col_name=["earliest_cr_line_Month","earliest_cr_line_Year"], pattern="-")), # split on '-'
("earliest_cr_line_Year_as_numeric", H2OColOp(op=H2OFrame.asnumeric, col="earliest_cr_line_Year", inplace=True)), # string -> double
# munge issue_d column in same way as earliest_cr_line column
("issue_date_split", H2OColOp(op=H2OFrame.strsplit, col="issue_d", inplace=False, new_col_name=["issue_d_Month", "issue_d_Year"], pattern="-")), # split on '-'
("issue_d_Year_as_numeric", H2OColOp(op=H2OFrame.asnumeric, col="issue_d_Year", inplace=True)), # string -> double
# do some munging of the emp_length column
("emp_length_rm_years", H2OColOp(op=H2OFrame.gsub, col="emp_length", inplace=True, pattern="([ ]*+[a-zA-Z].*)|(n/a)", replacement="")), # remove " year" and " years", also translate n/a to ""
("emp_length_trim", H2OColOp(op=H2OFrame.trim, col="emp_length", inplace=True)), # trim all the WS off
("emp_length_lt1_point5",H2OColOp(op=H2OFrame.gsub, col="emp_length", inplace=True, pattern="< 1", replacement="0.5")), # translate < 1 => 0.5
("emp_length_10plus", H2OColOp(op=H2OFrame.gsub, col="emp_length", inplace=True, pattern="10\\+", replacement="10")), # translate 10+ to 10
("emp_length_as_numeric",H2OColOp(op=H2OFrame.asnumeric, col="emp_length", inplace=True)), # string -> double
# compute credit length
("credit_length", H2OBinaryOp(op=H2OAssembly.minus, col="issue_d_Year",inplace=False, new_col_name="longest_credit_length",right=H2OCol("earliest_cr_line_Year")))
])
res = assembly.fit(data)
pyunit_utils.javamunge(assembly, "AssemblyMungingDemoPojo", test)
def __ml_train(X, extra_crispr_df, y, train_index, test_index):
logger.debug("Creating h2o working environment")
# ### Start H2O
# Start up a 1-node H2O cloud on your local machine, and allow it to use all CPU cores and up to 2GB of memory:
h2o.init(max_mem_size="2G")
h2o.remove_all()
logger.debug("Created h2o working environment successfully")
from h2o.estimators import H2ORandomForestEstimator
rf_crispr = H2ORandomForestEstimator(model_id="rf_crispr",
categorical_encoding="enum",
nfolds=5,
ntrees=30,
stopping_rounds=30,
score_each_iteration=True,
seed=10)
seq_data = X.iloc[:, :config.seq_len]
seq_data.columns = ['pos_' + str(i) for i in range(len(seq_data.columns))]
pre_h2o_df = pd.concat([seq_data, extra_crispr_df, y], axis=1)
h2o_crispr_df_train = h2o.H2OFrame(pre_h2o_df.loc[train_index, :])
h2o_crispr_df_test = h2o.H2OFrame(pre_h2o_df.loc[test_index, :])
logger.debug("Training machine learning model")
rf_crispr.train(x=h2o_crispr_df_train.col_names[:-1],
y=h2o_crispr_df_train.col_names[-1],
training_frame=h2o_crispr_df_train)
logger.debug("Trained successfully. Output feature importance")
feature_importance = rf_crispr._model_json['output'][
'variable_importances'].as_data_frame()[['variable', 'percentage']]
feature_importance.to_csv(config.feature_importance_path, index=False)
logger.debug("Predicting training data")
test_prediction_train = rf_crispr.predict(h2o_crispr_df_train[:-1])
performance = spearmanr(test_prediction_train.as_data_frame()['predict'],
h2o_crispr_df_train.as_data_frame()['log2fc'])[0]
logger.debug(
"spearman correlation coefficient for training dataset is: %f" %
performance)
logger.debug("Predicting test data")
test_prediction = rf_crispr.predict(h2o_crispr_df_test[:-1])
performance = spearmanr(test_prediction.as_data_frame()['predict'],
h2o_crispr_df_test.as_data_frame()['log2fc'])[0]
logger.debug("spearman correlation coefficient for test dataset is: %f" %
performance)
logger.debug("Saving model")
h2o.save_model(rf_crispr, config.ml_model_path)
logger.debug("Saved model to disk")
def init():
global sr, fr, share_cols
# <hack> regarding output to make h2o work in IDLE
class PseudoTTY(object):
def __init__(self, underlying):
underlying.encoding = 'cp437'
self.__underlying = underlying
def __getattr__(self, name):
return getattr(self.__underlying, name)
def isatty(self):
return True
import sys
sys.stdout = PseudoTTY(sys.stdout)
# </hack>
h2o.init(nthreads=-1, max_mem_size="58G")
h2o.remove_all()
init()
femq12 = pd.read_csv(
r"H:\Ashwin\dta\features\All_return_features_sample.csv")
# femq12['fold'] = (femq12['TIN_hash_byte']/32).astype(int)
fr = h2o.H2OFrame(python_obj=femq12)
print 'setting factors...'
fr = set_return_factors(fr)
fr = set_profile_factors(fr)
fr = set_match_factors(fr)
fr = set_transaction_factors(fr)
fr = set_purchasenetwork_factors(fr)
fr = set_salenetwork_factors(fr)
fr = set_downstream_factors(fr)
return fr
fr['Missing_SalesDSUnTaxProp'] = fr['Missing_SalesDSUnTaxProp'].asfactor()
fr['Missing_SalesDSCreditRatio'] = fr[
'Missing_SalesDSCreditRatio'].asfactor()
fr['Missing_SalesDSVatRatio'] = fr['Missing_SalesDSVatRatio'].asfactor()
fr['Missing_MaxSalesProp'] = fr['Missing_MaxSalesProp'].asfactor()
fr['Missing_MaxPurchaseProp'] = fr['Missing_MaxPurchaseProp'].asfactor()
fr['Missing_PurchaseDSUnTaxProp'] = fr[
'Missing_PurchaseDSUnTaxProp'].asfactor()
fr['Missing_PurchaseDSCreditRatio'] = fr[
'Missing_PurchaseDSCreditRatio'].asfactor()
fr['Missing_PurchaseDSVatRatio'] = fr[
'Missing_PurchaseDSVatRatio'].asfactor()
def __init__(self,
model,
X_test,
feature_names,
max_depth,
model_id='surrogate_mojo'):
self.model = model
self.X_test = np.array(X_test)
self.y_test = np.array(model.predict(X_test))
self.feature_names = feature_names
self.max_depth = max_depth
self.model_id = model_id
h2o.init(max_mem_size='2G') # start h2o
h2o.remove_all() # remove any existing data structures from h2o memory
def javamunge_assembly():
h2o.remove_all()
train = pyunit_utils.locate("bigdata/laptop/lending-club/LoanStats3a.csv")
test = pyunit_utils.locate("bigdata/laptop/lending-club/LoanStats3b.csv")
# lending-club munging assembly
print("Import and Parse data")
types = {"int_rate":"String", "revol_util":"String", "emp_length":"String"}
data = h2o.import_file(path=train, col_types=types)
test = h2o.import_file(path=test, col_types=types)
test = test[[1,5,19,23,45,66,99,590,8903,9999,10001,23892,23893,50123],:]
test = h2o.assign(test,"test")
assembly = H2OAssembly(
steps=[
# munge int_rate column in place
# strip %, trim ws, convert to double
("intrate_rm_junk_char", H2OColOp(op=H2OFrame.gsub, col="int_rate", inplace=True, pattern="%", replacement="")), # strip %
("intrate_trim_ws", H2OColOp(op=H2OFrame.trim, col="int_rate", inplace=True)), # trim ws
("intrate_as_numeric", H2OColOp(op=H2OFrame.asnumeric, col="int_rate", inplace=True)), # string -> double
# munge the revol_util in the same way as the int_rate column
("revol_rm_junk_char", H2OColOp(op=H2OFrame.gsub, col="revol_util", inplace=True, pattern="%", replacement="")), # strip %
("revol_trim_ws", H2OColOp(op=H2OFrame.trim, col="revol_util", inplace=True)), # trim ws
("revol_as_numeric", H2OColOp(op=H2OFrame.asnumeric, col="revol_util", inplace=True)), # string -> double
# munge earliest_cr_line column (mm-YYYY format)
# split into Month and Year columns
("earliest_cr_line_split", H2OColOp(H2OFrame.strsplit, col="earliest_cr_line", inplace=False, new_col_name=["earliest_cr_line_Month","earliest_cr_line_Year"], pattern="-")), # split on '-'
("earliest_cr_line_Year_as_numeric", H2OColOp(op=H2OFrame.asnumeric, col="earliest_cr_line_Year", inplace=True)), # string -> double
# munge issue_d column in same way as earliest_cr_line column
("issue_date_split", H2OColOp(op=H2OFrame.strsplit, col="issue_d", inplace=False, new_col_name=["issue_d_Month", "issue_d_Year"], pattern="-")), # split on '-'
("issue_d_Year_as_numeric", H2OColOp(op=H2OFrame.asnumeric, col="issue_d_Year", inplace=True)), # string -> double
# do some munging of the emp_length column
("emp_length_rm_years", H2OColOp(op=H2OFrame.gsub, col="emp_length", inplace=True, pattern="([ ]*+[a-zA-Z].*)|(n/a)", replacement="")), # remove " year" and " years", also translate n/a to ""
("emp_length_trim", H2OColOp(op=H2OFrame.trim, col="emp_length", inplace=True)), # trim all the WS off
("emp_length_lt1_point5",H2OColOp(op=H2OFrame.gsub, col="emp_length", inplace=True, pattern="< 1", replacement="0.5")), # translate < 1 => 0.5
("emp_length_10plus", H2OColOp(op=H2OFrame.gsub, col="emp_length", inplace=True, pattern="10\\+", replacement="10")), # translate 10+ to 10
("emp_length_as_numeric",H2OColOp(op=H2OFrame.asnumeric, col="emp_length", inplace=True)), # string -> double
# compute credit length
("credit_length", H2OBinaryOp(op=H2OAssembly.minus, col="issue_d_Year",inplace=False, new_col_name="longest_credit_length",right=H2OCol("earliest_cr_line_Year")))
])
res = assembly.fit(data)
pyunit_utils.javamunge(assembly, "AssemblyMungingDemoPojo", test)
def setup_and_train(param_enabled=None):
h2o.remove_all()
target, train, _, _ = prepare_data()
state = 'enabled' if param_enabled is True else 'disabled' if param_enabled is False else 'default'
if param_enabled is None:
aml = H2OAutoML(project_name='keep_cross_validation_predictions_'+state,
nfolds=nfolds, max_models=3, seed=1)
else:
aml = H2OAutoML(project_name='keep_cross_validation_predictions_'+state,
nfolds=nfolds, max_models=8, seed=1,
keep_cross_validation_predictions=param_enabled)
aml.train(y=target, training_frame=train)
# print(aml.leaderboard)
return aml
def run_test(sys_args, test_to_run):
global _IPYNB_
parse_args(sys_args)
h2o.init(ip=_H2O_IP_, port=_H2O_PORT_, strict_version_check=False)
h2o.log_and_echo("------------------------------------------------------------")
h2o.log_and_echo("")
h2o.log_and_echo("STARTING TEST: "+str(h2o.ou()))
h2o.log_and_echo("")
h2o.log_and_echo("------------------------------------------------------------")
# num_keys = h2o.store_size()
try:
if _IPYNB_: utils.ipy_notebook_exec(_IPYNB_, save_and_norun=False)
else: test_to_run()
finally:
h2o.remove_all()
def h2o_test_setup(sys_args):
h2o_py_dir = os.path.realpath(os.path.join(os.path.dirname(os.path.realpath(__file__)),".."))
h2o_docs_dir = os.path.realpath(os.path.join(os.path.dirname(os.path.realpath(__file__)),"..","..","h2o-docs"))
parse_args(sys_args)
sys.path.insert(1, h2o_py_dir)
import h2o
from tests import pyunit_utils, pydemo_utils, pybooklet_utils
for pkg in (pyunit_utils, pybooklet_utils):
setattr(pkg, '__on_hadoop__', _ON_HADOOP_)
setattr(pkg, '__hadoop_namenode__', _HADOOP_NAMENODE_)
setattr(pkg, '__test_name__', _TEST_NAME_)
setattr(pkg, '__results_dir__', _RESULTS_DIR_)
if _IS_PYUNIT_ or _IS_IPYNB_ or _IS_PYBOOKLET_ or _IS_PYDEMO_:
pass
else:
raise(EnvironmentError, "Unrecognized test type. Must be of type ipynb, pydemo, pyunit, or pybooklet, but got: "
"{0}".format(_TEST_NAME_))
print("[{0}] {1}\n".format(strftime("%Y-%m-%d %H:%M:%S", gmtime()), "Connect to h2o on IP: {0} PORT: {1}".format(_H2O_IP_, _H2O_PORT_)))
auth = None
if _LDAP_USER_NAME_ is not None and _LDAP_PASSWORD_ is not None:
auth = (_LDAP_USER_NAME_, _LDAP_PASSWORD_)
elif _KERB_PRINCIPAL_ is not None:
from h2o.auth import SpnegoAuth
auth = SpnegoAuth(service_principal=_KERB_PRINCIPAL_)
h2o.connect(ip=_H2O_IP_, port=_H2O_PORT_, verbose=False, auth=auth, **_H2O_EXTRA_CONNECT_ARGS_)
h2o.utils.config.H2OConfigReader.get_config()["general.allow_breaking_changes"] = True
#rest_log = os.path.join(_RESULTS_DIR_, "rest.log")
#h2o.start_logging(rest_log)
#print "[{0}] {1}\n".format(strftime("%Y-%m-%d %H:%M:%S", gmtime()), "Started rest logging in: {0}".format(rest_log))
h2o.log_and_echo("------------------------------------------------------------")
h2o.log_and_echo("")
h2o.log_and_echo("STARTING TEST: " + _TEST_NAME_)
h2o.log_and_echo("")
h2o.log_and_echo("------------------------------------------------------------")
h2o.remove_all()
if _IS_IPYNB_: pydemo_utils.ipy_notebook_exec(_TEST_NAME_)
elif _IS_PYUNIT_: pyunit_utils.pyunit_exec(_TEST_NAME_)
elif _IS_PYBOOKLET_: pybooklet_utils.pybooklet_exec(_TEST_NAME_)
elif _IS_PYDEMO_: pydemo_utils.pydemo_exec(_TEST_NAME_)
def sort():
df = h2o.create_frame(rows=10,
cols=3,
factors=10,
categorical_fraction=1.0/3,
time_fraction=1.0/3,
real_fraction=1.0/3,
real_range=100,
missing_fraction=0.0,
seed=123)
df1 = df.sort("C1")
assert df1[0,0] == 433225652950 # 1983-09-24 04:27:32
assert df1[9,0] == 1532907020199 # 2018-07-29 23:30:20
df2 = df.sort("C2")
assert df2[0,1] == "c1.l1"
assert df2[9,1] == "c1.l9"
h2o.remove_all()
def start_h2o(self, thread_count=-1, gb_ram_count=26):
"""Initializes a connection to H2O instance and clear it out, if needed.
:param thread_count: the number of threads that H2O may use or -1 if all available (Default value = -1)
:param gb_ram_count: the number of gigabytes of RAM that H2O may use (Default value = 26)
"""
h2o.init(nthreads=thread_count, max_mem_size=gb_ram_count)
# clear out cluster
return(h2o.remove_all())
def run_test(sys_args, test_to_run):
# import pkg_resources
# ver = pkg_resources.get_distribution("h2o").version
# print "H2O PYTHON PACKAGE VERSION: " + str(ver)
ip, port = sys_args[2].split(":")
h2o.init(ip,port,strict_version_check=False)
h2o.log_and_echo("------------------------------------------------------------")
h2o.log_and_echo("")
h2o.log_and_echo("STARTING TEST: "+str(h2o.ou()))
h2o.log_and_echo("")
h2o.log_and_echo("------------------------------------------------------------")
num_keys = h2o.store_size()
try:
if len(sys_args) > 3 and sys_args[3] == "--ipynb": utils.ipy_notebook_exec(sys_args[4],save_and_norun=False)
else: test_to_run(ip, port)
finally:
h2o.remove_all()
if h2o.keys_leaked(num_keys): print "Leaked Keys!"
def deepwater_lenet():
if not H2ODeepWaterEstimator.available(): return
frame = h2o.import_file(pyunit_utils.locate("bigdata/laptop/deepwater/imagenet/cat_dog_mouse.csv"))
print(frame.head(5))
model = H2ODeepWaterEstimator(epochs=100, learning_rate=1e-3, network='lenet', score_interval=0, train_samples_per_iteration=1000)
model.train(x=[0],y=1, training_frame=frame)
extracted = model.deepfeatures(frame, "pooling1_output")
#print(extracted.describe())
print(extracted.ncols)
assert extracted.ncols == 800, "extracted frame doesn't have 800 columns"
extracted = model.deepfeatures(frame, "activation2_output")
#print(extracted.describe())
print(extracted.ncols)
assert extracted.ncols == 500, "extracted frame doesn't have 500 columns"
h2o.remove_all()
def glrm_mojo():
h2o.remove_all()
NTESTROWS = 200 # number of test dataset rows
df = pyunit_utils.random_dataset("regression", seed=1234) # generate random dataset
train = df[NTESTROWS:, :]
test = df[:NTESTROWS, :]
x = df.names
transform_types = ["NONE", "STANDARDIZE", "NORMALIZE", "DEMEAN", "DESCALE"]
transformN = transform_types[randint(0, len(transform_types)-1)]
# build a GLRM model with random dataset generated earlier
glrmModel = H2OGeneralizedLowRankEstimator(k=3, transform=transformN, max_iterations=10, seed=1234)
glrmModel.train(x=x, training_frame=train)
glrmTrainFactor = h2o.get_frame(glrmModel._model_json['output']['representation_name'])
assert glrmTrainFactor.nrows==train.nrows, \
"X factor row number {0} should equal training row number {1}.".format(glrmTrainFactor.nrows, train.nrows)
save_GLRM_mojo(glrmModel) # ave mojo model
MOJONAME = pyunit_utils.getMojoName(glrmModel._id)
TMPDIR = os.path.normpath(os.path.join(os.path.dirname(os.path.realpath('__file__')), "..", "results", MOJONAME))
h2o.download_csv(test[x], os.path.join(TMPDIR, 'in.csv')) # save test file, h2o predict/mojo use same file
pred_h2o, pred_mojo = pyunit_utils.mojo_predict(glrmModel, TMPDIR, MOJONAME, glrmReconstruct=True) # save mojo predict
h2o.save_model(glrmModel, TMPDIR) # save GLRM model
glrmModel2 = h2o.load_model(os.path.join(TMPDIR,MOJONAME))
predict_model = glrmModel2.predict(test)
for col in range(pred_h2o.ncols):
if pred_h2o[col].isfactor():
pred_h2o[col] = pred_h2o[col].asnumeric()
predict_model[col] = predict_model[col].asnumeric()
print("Comparing mojo predict and h2o predict...")
pyunit_utils.compare_frames_local(pred_h2o, pred_mojo, 1, tol=1e-10)
print("Comparing mojo predict and h2o predict from saved model...")
pyunit_utils.compare_frames_local(pred_mojo, predict_model, 1, tol=1e-10)
frameID, mojoXFactor = pyunit_utils.mojo_predict(glrmModel, TMPDIR, MOJONAME, glrmReconstruct=False) # save mojo XFactor
glrmTestFactor = h2o.get_frame("GLRMLoading_"+frameID) # store the x Factor for new test dataset
print("Comparing mojo x Factor and model x Factor ...")
pyunit_utils.compare_frames_local(glrmTestFactor, mojoXFactor, 1, tol=1e-10)
def h2o_test_setup(sys_args):
h2o_py_dir = os.path.realpath(os.path.join(os.path.dirname(os.path.realpath(__file__)),".."))
h2o_docs_dir = os.path.realpath(os.path.join(os.path.dirname(os.path.realpath(__file__)),"..","..","h2o-docs"))
parse_args(sys_args)
sys.path.insert(1, h2o_py_dir)
import h2o
from tests import pyunit_utils, pydemo_utils, pybooklet_utils
set_pyunit_pkg_attrs(pyunit_utils)
set_pybooklet_pkg_attrs(pybooklet_utils)
if _IS_PYUNIT_ or _IS_IPYNB_ or _IS_PYBOOKLET_:
pass
elif _IS_PYDEMO_:
raise(NotImplementedError, "pydemos are not supported at this time")
else:
raise(EnvironmentError, "Unrecognized test type. Must be of type ipynb, pydemo, pyunit, or pybooklet, but got: "
"{0}".format(_TEST_NAME_))
print "[{0}] {1}\n".format(strftime("%Y-%m-%d %H:%M:%S", gmtime()), "Connect to h2o on IP: {0} PORT: {1}"
"".format(_H2O_IP_, _H2O_PORT_))
h2o.init(ip=_H2O_IP_, port=_H2O_PORT_, strict_version_check=False)
#rest_log = os.path.join(_RESULTS_DIR_, "rest.log")
#h2o.start_logging(rest_log)
#print "[{0}] {1}\n".format(strftime("%Y-%m-%d %H:%M:%S", gmtime()), "Started rest logging in: {0}".format(rest_log))
h2o.log_and_echo("------------------------------------------------------------")
h2o.log_and_echo("")
h2o.log_and_echo("STARTING TEST: " + _TEST_NAME_)
h2o.log_and_echo("")
h2o.log_and_echo("------------------------------------------------------------")
h2o.remove_all()
if _IS_IPYNB_: pydemo_utils.ipy_notebook_exec(_TEST_NAME_)
elif _IS_PYUNIT_: pyunit_utils.pyunit_exec(_TEST_NAME_)
elif _IS_PYBOOKLET_: pybooklet_utils.pybooklet_exec(_TEST_NAME_)
def glm_binomial_mojo_pojo():
h2o.remove_all()
NTESTROWS = 200 # number of test dataset rows
PROBLEM="binomial"
params = set_params() # set deeplearning model parameters
df = pyunit_utils.random_dataset(PROBLEM) # generate random dataset
train = df[NTESTROWS:, :]
test = df[:NTESTROWS, :]
x = list(set(df.names) - {"response"})
glmBinomialModel = pyunit_utils.build_save_model_GLM(params, x, train, "response") # build and save mojo model
MOJONAME = pyunit_utils.getMojoName(glmBinomialModel._id)
TMPDIR = os.path.normpath(os.path.join(os.path.dirname(os.path.realpath('__file__')), "..", "results", MOJONAME))
h2o.download_csv(test[x], os.path.join(TMPDIR, 'in.csv')) # save test file, h2o predict/mojo use same file
pred_h2o, pred_mojo = pyunit_utils.mojo_predict(glmBinomialModel, TMPDIR, MOJONAME) # load model and perform predict
h2o.download_csv(pred_h2o, os.path.join(TMPDIR, "h2oPred.csv"))
pred_pojo = pyunit_utils.pojo_predict(glmBinomialModel, TMPDIR, MOJONAME)
print("Comparing mojo predict and h2o predict...")
pyunit_utils.compare_frames_local(pred_h2o, pred_mojo, 0.1, tol=1e-10) # make sure operation sequence is preserved from Tomk h2o.save_model(glmOrdinalModel, path=TMPDIR, force=True) # save model for debugging
print("Comparing pojo predict and h2o predict...")
pyunit_utils.compare_frames_local(pred_mojo, pred_pojo, 0.1, tol=1e-10)
def pca_wideDataset_rotterdam():
h2o.remove_all()
print("Importing Rotterdam.csv data...")
rotterdamH2O = h2o.upload_file(pyunit_utils.locate("bigdata/laptop/jira/rotterdam.csv.zip"))
y = set(["relapse"])
x = list(set(rotterdamH2O.names)-y)
transform_types = ["NONE", "STANDARDIZE", "NORMALIZE", "DEMEAN", "DESCALE"]
transformN = transform_types[randint(0, len(transform_types)-1)]
print("transform used on dataset is {0}.\n".format(transformN))
buildModel = [False, False, False]
buildModel[randint(0, len(buildModel)-1)] = True
expNum = 0
if (buildModel[expNum]):
# special test with GLRM. Need use_all_levels to be true
print("------ Testing GLRM PCA --------")
gramSVD = H2OPCA(k=8, impute_missing=True, transform=transformN, seed=12345, use_all_factor_levels=True)
gramSVD.train(x=x, training_frame=rotterdamH2O)
glrmPCA = H2OGeneralizedLowRankEstimator(k=8, transform=transformN, seed=12345, init="Random",
max_iterations=10, recover_svd=True, regularization_x="None",
regularization_y="None")
glrmPCA.train(x=x, training_frame=rotterdamH2O)
# compare singular values and stuff with GramSVD
print("@@@@@@ Comparing eigenvectors between GramSVD and GLRM...\n")
print("@@@@@@ Comparing eigenvalues between GramSVD and GLRM...\n")
pyunit_utils.assert_H2OTwoDimTable_equal(gramSVD._model_json["output"]["importance"],
glrmPCA._model_json["output"]["importance"],
["Standard deviation", "Cumulative Proportion", "Cumulative Proportion"],
tolerance=1, check_all=False)
# compare singular vectors
pyunit_utils.assert_H2OTwoDimTable_equal(gramSVD._model_json["output"]["eigenvectors"],
glrmPCA._model_json["output"]["eigenvectors"],
glrmPCA._model_json["output"]["names"], tolerance=1e-6,
check_sign=True, check_all=False)
h2o.remove(gramSVD)
h2o.remove(glrmPCA)
expNum=expNum+1
if (buildModel[expNum]):
print("------ Testing Power PCA --------")
gramSVD = H2OPCA(k=8, impute_missing=True, transform=transformN, seed=12345)
gramSVD.train(x=x, training_frame=rotterdamH2O)
powerPCA = H2OPCA(k=8, impute_missing=True, transform=transformN, pca_method="Power", seed=12345) # power
powerPCA.train(x=x, training_frame=rotterdamH2O)
# compare singular values and stuff with GramSVD
print("@@@@@@ Comparing eigenvalues between GramSVD and Power...\n")
pyunit_utils.assert_H2OTwoDimTable_equal(gramSVD._model_json["output"]["importance"],
powerPCA._model_json["output"]["importance"],
["Standard deviation", "Cumulative Proportion", "Cumulative Proportion"],
tolerance=1e-6, check_all=False)
print("@@@@@@ Comparing eigenvectors between GramSVD and Power...\n")
# compare singular vectors
pyunit_utils.assert_H2OTwoDimTable_equal(gramSVD._model_json["output"]["eigenvectors"],
powerPCA._model_json["output"]["eigenvectors"],
powerPCA._model_json["output"]["names"], tolerance=1e-6, check_sign=True,
check_all=False)
expNum=expNum+1
if (buildModel[expNum]):
print("------ Testing Randomized PCA --------")
gramSVD = H2OPCA(k=8, impute_missing=True, transform=transformN, seed=12345)
gramSVD.train(x=x, training_frame=rotterdamH2O)
randomizedPCA = H2OPCA(k=8, impute_missing=True, transform=transformN, pca_method="Randomized", seed=12345,
max_iterations=5) # power
randomizedPCA.train(x=x, training_frame=rotterdamH2O)
# compare singular values and stuff with GramSVD
print("@@@@@@ Comparing eigenvalues between GramSVD and Randomized...\n")
pyunit_utils.assert_H2OTwoDimTable_equal(gramSVD._model_json["output"]["importance"],
randomizedPCA._model_json["output"]["importance"],
["Standard deviation", "Cumulative Proportion", "Cumulative Proportion"],
tolerance=1e-1, check_all=False)
print("@@@@@@ Comparing eigenvectors between GramSVD and Power...\n")
# compare singular vectors
pyunit_utils.assert_H2OTwoDimTable_equal(gramSVD._model_json["output"]["eigenvectors"],
randomizedPCA._model_json["output"]["eigenvectors"],
randomizedPCA._model_json["output"]["names"], tolerance=1e-6,
check_sign=True, check_all=False)
h2o.remove_all()
def setup_dataset():
h2o.remove_all()
train = h2o.import_file(path=pyunit_utils.locate("smalldata/iris/iris_wheader.csv"))
return train
# coding: utf-8
# In[ ]:
import h2o
import numpy as np
import matplotlib.patches as mpatches
import matplotlib.pyplot as plt
from h2o.estimators.glrm import H2OGeneralizedLowRankEstimator
import os
# In[ ]:
h2o.init()
h2o.remove_all() # Clean slate - just in case the cluster was already running
# In[ ]:
from h2o.h2o import _locate # private function. used to find files within h2o git project directory.
# Import walking gait data
gait = h2o.import_file(path=os.path.realpath("../data/subject01_walk1.csv"))
gait.describe()
# In[ ]:
# Plot first row of data on x- vs. y-coordinate features
gait_row = gait[1,:].drop("Time")
def s3timings(ip, port):
t = time.time()
# connect to cluster
h2o.init(ip, port)
# defining timers
air_run = timeit.Timer(stmt = 'h2o.import_frame("s3n://h2o-airlines-unpacked/allyears.1987.2013.csv")',
setup = 'import h2o')
bigx_run = timeit.Timer(stmt = 'h2o.import_frame("s3://h2o-public-test-data/bigdata/server/flow-tests/BigCross.data")',
setup = 'import h2o')
higg_run = timeit.Timer(stmt = 'h2o.import_frame("s3://h2o-public-test-data/bigdata/server/HIGGS.csv")',
setup = 'import h2o')
citi_run = timeit.Timer(stmt = 'h2o.import_frame(path = big_citi)',
setup = 'import h2o;\
big_citi = ["s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2013-07.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2013-08.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2013-09.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2013-10.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2013-11.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2013-12.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2014-01.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2014-02.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2014-03.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2014-04.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2014-05.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2014-06.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2014-07.csv",\
"s3://h2o-public-test-data/bigdata/laptop/citibike-nyc/2014-08.csv"]')
mils_run = timeit.Timer(stmt = 'h2o.import_frame(path = mill_songs)',
setup = 'import h2o;\
mill_songs = ["s3://h2o-public-test-data/bigdata/server/milsongs/milsongs-test.csv",\
"s3://h2o-public-test-data/bigdata/server/milsongs/milsongs-train.csv"]')
cup_run = timeit.Timer(stmt = 'h2o.import_frame(path = cup98)',
setup = 'import h2o;\
cup98 = ["s3://h2o-public-test-data/bigdata/laptop/usecases/cup98LRN_z.csv",\
"s3://h2o-public-test-data/bigdata/laptop/usecases/cup98VAL_z.csv"]')
mnist_run = timeit.Timer(stmt = 'h2o.import_frame(path = mnist)',
setup = 'import h2o;\
mnist = ["s3://h2o-public-test-data/bigdata/laptop/mnist/test.csv.gz",\
"s3://h2o-public-test-data/bigdata/laptop/mnist/train.csv.gz"]')
arc_run = timeit.Timer(stmt = 'h2o.import_frame(path = arcene)',
setup = 'import h2o;\
arcene = ["s3://h2o-public-test-data/smalldata/arcene/arcene_test.data",\
"s3://h2o-public-test-data/smalldata/arcene/arcene_train.data",\
"s3://h2o-public-test-data/smalldata/arcene/arcene_valid.data"]')
# Running with timers
air_first = air_run.timeit(number=1)
bigx_first = bigx_run.timeit(number=1)
higg_first = higg_run.timeit(number=1)
citi_first = citi_run.timeit(number=1)
mils_first = mils_run.timeit(number=1)
cup_first = cup_run.timeit(number=1)
mnist_first = mnist_run.timeit(number=1)
arc_first = arc_run.timeit(number=1)
# Clear kvstore and run again
s = time.time()
h2o.remove_all()
print "Elapsed Time for RemoveAll: " + str(time.time() - s) + " (s)."
air_second = air_run.timeit(number=1)
bigx_second = bigx_run.timeit(number=1)
higg_second = higg_run.timeit(number=1)
citi_second = citi_run.timeit(number=1)
mils_second = mils_run.timeit(number=1)
cup_second = cup_run.timeit(number=1)
mnist_second = mnist_run.timeit(number=1)
arc_second = arc_run.timeit(number=1)
print("Airlines: " + str(air_first) + " vs " + str(air_second))
print("BigCross: " + str(bigx_first) + " vs " + str(bigx_second))
print("Higgs: " + str(higg_first) + " vs " + str(higg_second))
print("Citi_bikes: " + str(citi_first) + " vs " + str(citi_second))
print("Million Songs: " + str(mils_first) + " vs " + str(mils_second))
print("KDD Cup98: " + str(cup_first) + " vs " + str(cup_second))
print("Mnist: " + str(mnist_first) + " vs " + str(mnist_second))
print("Arcene: " + str(arc_first) + " vs " + str(arc_second))
s = time.time()
h2o.remove_all()
print "Elapsed Time for RemoveAll: " + str(time.time() - s) + " (s)."
print "Exiting scope... Test elapsed time: " + str(time.time() - t) + " (s)."
# # Load the H2O Python module. # In[ ]: import h2o import os # ### Start H2O # Start up a 1-node H2O cloud on your local machine, and allow it to use all CPU cores and up to 2GB of memory: # In[ ]: h2o.init(max_mem_size_GB = 2) #uses all cores by default h2o.remove_all() #clean slate, in case cluster was already running # To learn more about the h2o package itself, we can use Python's builtin help() function. # In[ ]: help(h2o) # help() can be used on H2O functions and models. Jupyter's builtin shift-tab functionality also works # In[ ]: from h2o.estimators.gbm import H2OGradientBoostingEstimator from h2o.estimators.random_forest import H2ORandomForestEstimator