# mapping from trial identifiers to corresponding input filenames
trial_input_fnames = id2filenames(trial_dir, "input", trial_ids)
# mapping from trial identifiers to corresponding gold standard filenames
trial_gs_fnames = id2filenames(trial_dir, "gs", trial_ids)
# mappings, ids and path for the 2014 test data
test_dir = join(data_dir, "STS2014-test")
test_ids = trial_ids
test_input_fnames = id2filenames(test_dir, "input", test_ids)
test_gs_fnames = id2filenames(test_dir, "gs", test_ids)
# mapping from test dataset identifiers and feature names
# to the corresponding feature files
test_feat_fnames = map_id_to_feat_files(os.path.join(feat_dir, 'STS2014-test'),
test_ids)
def read_test_data(ids, features=[], convert_nan=True):
"""
Create feature vectors and labels for given dataset identifiers and
features from STS13 test data
"""
return read_data(ids,
test_feat_fnames,
test_gs_fnames,
features=features,
convert_nan=convert_nan)
def read_blind_test_data(ids, features=[], convert_nan=True):
define dirs, ids and filenames for STS13 test data
"""
import os
from os.path import join
from ntnu.io import read_data, map_id_to_feat_files, feat_dir
from sts.io import data_dir, id2filenames
# directory containing original STS 2103 test files
test_dir = join(data_dir, "STS2013-test")
# identifiers for different categories of test data
test_ids = "FNWN", "headlines", "OnWN", "SMT"
# mapping from test identifiers to corresponding input filenames
test_input_fnames = id2filenames(test_dir, "input", test_ids)
# mapping from test identifiers to corresponding gold standard filenames
test_gs_fnames = id2filenames(test_dir, "gs", test_ids)
test_feat_fnames = map_id_to_feat_files(os.path.join(feat_dir, 'STS2013-test'), test_ids)
def read_test_data(ids, features=[], convert_nan=True):
"""
Create feature vectors and labels for given dataset identifiers and
features from STS12 test data
"""
return read_data(ids, test_feat_fnames, test_gs_fnames,
features=features, convert_nan=convert_nan )
"""
define dirs and filenames of features for STS12 data
"""
from os.path import join
from sts.sts12 import train_ids, test_ids, train_gs_fnames, test_gs_fnames
from ntnu.io import feat_dir, read_data, map_id_to_feat_files
# top directory containing train feature files
train_dir = join(feat_dir, "STS2012-train")
# mapping from train dataset identifiers and feature names
# to the corresponding feature files
train_feat_fnames = map_id_to_feat_files(train_dir, train_ids)
def read_train_data(ids, features=[], convert_nan=True):
"""
Create feature vectors and labels for given dataset identifiers and
features from STS12 train data
"""
return read_data(ids,
train_feat_fnames,
train_gs_fnames,
features=features,
convert_nan=convert_nan)
# top directory containing test feature files
define dirs and filenames of features for STS14 data
"""
from os.path import join
from sts.sts14 import trial_ids, trial_gs_fnames
from ntnu.io import feat_dir, read_data, read_blind_data, map_id_to_feat_files
# top directory containing trial feature files
trial_dir = join(feat_dir, "STS2014-trial")
# mapping from test dataset identifiers and feature names
# to the corresponding feature files
trial_feat_fnames = map_id_to_feat_files(trial_dir, trial_ids)
def read_trial_data(ids, features=[], convert_nan=True):
"""
Create feature vectors and labels for given dataset identifiers and
features from STS14 trial data
"""
return read_data(ids, trial_feat_fnames, trial_gs_fnames,
features=features, convert_nan=convert_nan)
def read_blind_trial_data(ids, features=[], convert_nan=True):
return read_blind_data(ids, trial_feat_fnames, features=features,
convert_nan=convert_nan )
"""
from os.path import join
from sts.sts12 import train_ids, test_ids, train_gs_fnames, test_gs_fnames
from ntnu.io import feat_dir, read_data, map_id_to_feat_files
# top directory containing train feature files
train_dir = join(feat_dir, "STS2012-train")
# mapping from train dataset identifiers and feature names
# to the corresponding feature files
train_feat_fnames = map_id_to_feat_files(train_dir, train_ids)
def read_train_data(ids, features=[], convert_nan=True):
"""
Create feature vectors and labels for given dataset identifiers and
features from STS12 train data
"""
return read_data(ids, train_feat_fnames, train_gs_fnames,
features=features, convert_nan=convert_nan)
# top directory containing test feature files
test_dir = join(feat_dir, "STS2012-test")
# mapping from test dataset identifiers and feature names
"""
define dirs and filenames of features for STS14 data
"""
from os.path import join
from sts.sts14 import trial_ids, trial_gs_fnames
from ntnu.io import feat_dir, read_data, read_blind_data, map_id_to_feat_files
# top directory containing trial feature files
trial_dir = join(feat_dir, "STS2014-trial")
# mapping from test dataset identifiers and feature names
# to the corresponding feature files
trial_feat_fnames = map_id_to_feat_files(trial_dir, trial_ids)
def read_trial_data(ids, features=[], convert_nan=True):
"""
Create feature vectors and labels for given dataset identifiers and
features from STS14 trial data
"""
return read_data(ids,
trial_feat_fnames,
trial_gs_fnames,
features=features,
convert_nan=convert_nan)
def read_blind_trial_data(ids, features=[], convert_nan=True):
define dirs and filenames of features for STS13 data
"""
from os.path import join
from sts.sts13 import test_ids, test_gs_fnames
from ntnu.io import feat_dir, read_data, read_blind_data, map_id_to_feat_files
# top directory containing test feature files
test_dir = join(feat_dir, "STS2013-test")
# mapping from test dataset identifiers and feature names
# to the corresponding feature files
test_feat_fnames = map_id_to_feat_files(test_dir, test_ids)
def read_test_data(ids, features=[], convert_nan=True):
"""
Create feature vectors and labels for given dataset identifiers and
features from STS13 test data
"""
return read_data(ids, test_feat_fnames, test_gs_fnames,
features=features, convert_nan=convert_nan)
def read_blind_test_data(ids, features=[], convert_nan=True):
return read_blind_data(ids, test_feat_fnames, features=features,
convert_nan=convert_nan )
