def get_features_array(self,
prop_graphs,
micro_features,
macro_features,
news_source=None,
label=None,
file_dir="data/features",
use_cache=False):
function_refs = []
file_name = self.get_dump_file_name(news_source, micro_features,
macro_features, label, file_dir)
data_file = Path(file_name)
if use_cache and data_file.is_file():
return pickle.load(open(file_name, "rb"))
if micro_features:
function_refs.extend(self.get_micro_feature_method_references())
if len(function_refs) == 0:
return None
all_features = []
for idx in range(len(function_refs)):
features_set = get_sample_feature_value(prop_graphs,
function_refs[idx])
all_features.append(features_set)
feature_array = np.transpose(get_numpy_array(all_features))
pickle.dump(feature_array, open(file_name, "wb"))
return feature_array
def get_all_linguistic_features(news_graphs, micro_features, macro_features):
all_features = []
if macro_features:
retweet_function_references = []
for function_reference in retweet_function_references:
features_set = get_stats_for_features(news_graphs,
function_reference,
print=False,
feature_name=None)
all_features.append(features_set)
if micro_features:
reply_function_references = [
get_reply_nodes_average_sentiment,
get_first_reply_nodes_average_sentiment,
get_deepest_cascade_reply_nodes_avg_sentiment,
get_deepest_cascade_first_level_reply_sentiment
]
for function_reference in reply_function_references:
features_set = get_stats_for_features(news_graphs,
function_reference,
print=True,
feature_name=None)
all_features.append(features_set)
return np.transpose(get_numpy_array(all_features))
def get_features_array(self, prop_graphs, micro_features, macro_features, news_source=None, label=None,
file_dir="data/features", use_cache=False):
all_features = []
file_name = self.get_dump_file_name(news_source, micro_features, macro_features, label, file_dir)
data_file = Path(file_name)
if use_cache and data_file.is_file():
return pickle.load(open(file_name, "rb"))
if micro_features:
target_edge_type = REPLY_EDGE
reply_function_references = self.get_micro_feature_method_references()
for function_ref in reply_function_references:
features = function_ref(prop_graphs, target_edge_type)
all_features.append(features)
if macro_features:
target_edge_type = RETWEET_EDGE
retweet_function_references = self.get_macro_feature_method_references()
for function_ref in retweet_function_references:
features = function_ref(prop_graphs, target_edge_type)
all_features.append(features)
feature_array = np.transpose(get_numpy_array(all_features))
pickle.dump(feature_array, open(file_name, "wb"))
return feature_array
def get_all_temporal_features(prop_graphs, micro_features, macro_features):
macro_features_functions = [
get_average_time_between_post_tweets,
get_time_diff_first_last_post_tweet,
get_time_diff_first_post_last_retweet,
get_time_diff_first_post_first_retweet, get_avg_time_between_retweets,
get_avg_retweet_time_deepest_cascade,
get_time_diff_post_time_last_retweet_time_deepest_cascade
]
micro_features_functions = [
get_avg_time_between_replies, get_time_diff_first_post_last_reply,
get_time_diff_post_time_last_reply_time_deepest_cascade
]
function_refs = []
if macro_features:
function_refs.extend(macro_features_functions)
if micro_features:
function_refs.extend(micro_features_functions)
all_features = []
for function_reference in function_refs:
features_set = get_stats_for_features(prop_graphs,
function_reference,
print=False,
feature_name=None)
all_features.append(features_set)
return np.transpose(get_numpy_array(all_features))
def get_all_structural_features(news_graphs, micro_features, macro_features):
all_features = []
target_edge_type = RETWEET_EDGE
if macro_features:
retweet_function_references = [get_tree_heights, get_prop_graphs_node_counts, get_prop_graps_cascade_num,
get_max_outdegrees, get_num_of_cascades_with_retweets,
get_fraction_of_cascades_with_retweets]
for function_ref in retweet_function_references:
features = function_ref(news_graphs, target_edge_type)
all_features.append(features)
if micro_features:
target_edge_type = REPLY_EDGE
reply_function_references = [get_tree_heights, get_prop_graphs_node_counts, get_max_outdegrees]
for function_ref in reply_function_references:
features = function_ref(news_graphs, target_edge_type)
all_features.append(features)
return np.transpose(get_numpy_array(all_features))