def tabulate_genre_dist(y,normalize_to_level=1):
"""
Takes a vector of lists of genres and tabulate the distribution of each genre in the dataset
Prints to console a ascii table of the distribution with count and percentage
:param y:
:return:
"""
genre_to_count=coll.Counter()
y=genre_normalizer(y,normalize_to_level)
if np.issubdtype(y.dtype,np.str):
genre_to_count.update(y)
total=y.shape[0]
else:
genre_to_count.update((i for i in itertools.chain(*y)))
total=sum((len(i) for i in y))
headers=["genre","count","percent"]
data=[headers]
data.extend([[k,v,round(v/total,2)] for k,v in genre_to_count.items()])
print(tabulate(data,headers="firstrow"))
def extract_meta_data(reference_db_cls,db_cls):
"""
For selected webpages in URLToGenre:
Extract meta data descriptions(name=description) and keywords and form bag of words representation with it.
Store it into a database
:return: None
"""
comp_logger.info("Extracting from the database {}, putting into {}".format(reference_db_cls,db_cls))
bow_transformer=BagOfWords()
not_found_data=0
for c,ref_object in enumerate(reference_db_cls.objects.no_cache()):
c%10000==0 and comp_logger.info("Done with {} MetaDatas".format(c))
url=ref_object.url
ref_index=ref_object.ref_index
short_genres=genre_normalizer(ref_object.short_genres,dim=1)
page=URLToGenre.objects(url=url).only("page")[0].page
page_soup=BeautifulSoup(page,"html.parser")
contents=[]
try:
for meta_data_desc in page_soup.find_all("meta",{"name":"description"}):
contents.append(meta_data_desc["content"])
for meta_data_desc in page_soup.find_all("meta",{"name":"Description"}):
contents.append(meta_data_desc["content"])
for meta_data_desc in page_soup.find_all("meta",{"name":"keywords"}):
contents.append(meta_data_desc["content"])
contents=" ".join(contents if contents else "")
#meta_bow=bow_transformer.get_word_count(contents) if contents and contents.strip() else {}
if not len(contents):
not_found_data+=1
except (KeyError,AttributeError,ValueError):
not_found_data+=1
meta_bow={}
#store into db
#db_cls(ref_index=ref_index,attr_map=meta_bow,short_genres=short_genres).save()
comp_logger.info("The MetaData does not exists in {} instances".format(not_found_data))
def _generate_mixed_effect_matrix(X_path,y_path,feat_selector):
"""
Converts X to a COO Matrix of Mixed effect matrix
:param X_path:
:param y_path:
:param feat_selector:
:return:
"""
mixed_effect_logger.debug("Flattening")
#Reduce the column count
X,y,_=flatten_set(*random_pick_samples(unpickle_obj(X_path),genre_normalizer(unpickle_obj(y_path))))
feat_selector.fit(X,y)
mixed_effect_logger.debug("Final size of X: {} y:{}".format(X.shape,y.shape))
#Get the column selector, indices
vocab_selector=feat_selector.get_support(True)
num_vocab=vocab_selector.shape[0]
vstack_list=[0]*X.shape[0]
for ind,X_row in enumerate(X):
ind % 10==0 and mixed_effect_logger.info("Done with {}".format(ind))
row=np.zeros((1,num_vocab**2))
select_col=X_row[0,vocab_selector].toarray() #convert to dense rep.
#Compare each index to each row. Record the minimum as cooccurence
for col_ind in range(0,select_col.shape[1]):
if not select_col[0,col_ind]:
continue
cmp=np.full((1,select_col.shape[1]),fill_value=select_col[0,col_ind])
select_col=np.minimum(select_col,cmp)
row[0,col_ind*num_vocab:(col_ind+1)*num_vocab]=select_col
vstack_list[ind]=lil_matrix(row)
del row,select_col
return vstack(vstack_list).tocoo()
def extract_title(reference_db_cls,db_cls):
"""
Extract title from some webpage in URLToGenre and save it to the db_cls database
reference db's object must have url and ref_index attributes
:param db_cls:
:return:
"""
comp_logger.info("Extracting from the database {}, putting into {}".format(reference_db_cls,db_cls))
bow_transformer=BagOfWords()
title_not_exists=0
for c,ref_object in enumerate(reference_db_cls.objects.no_cache()):
c%10==0 and comp_logger.info("Done with {} titles".format(c))
url=ref_object.url
ref_index=ref_object.ref_index
short_genres=genre_normalizer(ref_object.short_genres,dim=1)
page=URLToGenre.objects(url=url).only("page")[0].page
page_soup=BeautifulSoup(page,"html.parser")
try:
title=page_soup.title.string
#bag of word
#title_bow=bow_transformer.get_word_count(title) if title and title.strip() else {}
except (AttributeError,ValueError):
title_not_exists+=1
title_bow={}
#store into db
#db_cls(ref_index=ref_index,attr_map=title_bow,short_genres=short_genres).save()
comp_logger.info("The title does not exists in {} instances".format(title_not_exists))
