cumulative=False,
normed=True,
label=['pos', 'neg'])
ax2.legend()
plt.show()
if __name__ == '__main__':
yc = ConfigParser.ConfigParser()
yc.read(CONFIG)
sfx = yc.get('Path', 'yep_data_class')
st = Stat(user_base = Utility.load_user_filtered(L_USER_F + sfx), # specify user base
user_label = Utility.load_user_label(R_USER_LABEL + sfx)) # specify user label
# st.co_review_num_dist(Utility.load_user_buz_rating(R_USER_BUZ_RATING + sfx))
# st.co_review_rating_dev_dist(Utility.load_user_buz_rating(R_USER_BUZ_RATING + sfx))
# st.knn_sim_distribution(np.iinfo(np.int64).max, Yelp.load_sim_mat(SM_COMMON_FRIEND + sfx))
# st.knn_sim_distribution(10, Yelp.load_sim_mat(SM_COMMON_FRIEND + sfx))
# st.knn_sim_distribution(5, Yelp.load_sim_mat(SM_COMMON_FRIEND + sfx))
# st.knn_sim_distribution(2, Yelp.load_sim_mat(SM_COMMON_FRIEND + sfx))
# st.knn_sim_distribution(1, Yelp.load_sim_mat(SM_COMMON_FRIEND + sfx))
# st.knn_sim_distribution(np.iinfo(np.int64).max, Yelp.load_sim_mat(SM_RATING_DEVIATION + sfx))
# st.knn_sim_distribution(np.iinfo(np.int64).max, Yelp.load_sim_mat(SM_COREVIEW_RATIO + sfx))
# st.knn_sim_distribution(1, Yelp.load_sim_mat(SM_COREVIEW_RATIO + sfx))
return sm
@staticmethod
def show_mat(sm):
# print sm
plt.spy(sm, marker='.', precision=0.1, markersize=5)
plt.show()
if __name__ == '__main__':
yc = ConfigParser.ConfigParser()
yc.read(CONFIG)
sfx = yc.get('Path', 'yep_data_class') # indicate which yelp dataset is used - _boston or _SanFrancisco
y = Yelp(sfx = sfx, # specify dataset class
user_base = Utility.load_user_filtered(L_USER_F + sfx)) # specify user base
#===============================================================================
# Generate Simialrity Matrix
#===============================================================================
# 1) Common Friend
# user_friend = Utility.load_user_friends(R_USER_FRIEND + sfx)
# y.gen_sim_mat(Yelp.pf_common_friend, SM_COMMON_FRIEND + sfx , (user_friend,))
# 2) Rating Deviation
# user_buz_rating = Utility.load_user_buz_rating(R_USER_BUZ_RATING + sfx)
# y.gen_sim_mat(Yelp.pf_rating_deviation, SM_RATING_DEVIATION + sfx , (user_buz_rating,))
# 3) Co-review ratio
# user_buz_rating = Utility.load_user_buz_rating(R_USER_BUZ_RATING + sfx)
# y.gen_sim_mat(Yelp.pf_co_review_ratio, SM_COREVIEW_RATIO + sfx, (user_buz_rating,))