def predict(trainer,ent_id,creator='monk'):#How? ent = ms.load_entity(ent_id) print ent.desc # return sign0(trainer.pandas[0].predict(ent)) score=trainer.pandas[0].predict(ent) print score return score
def get_recommended_place(collection_name,skip_num=0,request_num=5): #print collection_name ents = ms.load_entities(collectionName = collection_name) user_turtle = 'likeTravel' trainer = train(user_turtle,'monk') rank=[] for ent in ents: rank.append((ent._id,predict(trainer,ent._id))) sorted_by_score = sorted(rank, key=lambda tup: tup[1]) rst = [] for r in sorted_by_score: ent = ms.load_entity(r[0]) e = {} # e['place_id'] = ent.place_id e['_id'] = ent._id e['img_url'] = ent.img_url e['desc'] = ent.desc rst.append(e) # print len(ents) return rst[skip_num:skip_num+request_num]
[ent.save() for ent in ents] likeTS = ms.yaml2json('turtle_scripts/turtle_like.yml') # print likeTS likeT = ms.create_turtle(likeTS) likeT.save() ent = ents[0] # print ents[0].generic() ent._setattr('likeTravel', 'Y') ms.crane.entityStore.save_one(ent) ms.add_data('likeTravel', 'monk', str(ents[0]._id)) print likeT.tigress.p print likeT.pandas[0].mantis.data likeT.tigress.defaulting=True likeT.save() likeT = ms.load_turtle('likeTravel','monk') likeT.train() for i in ents: ent = ms.load_entity(i._id) print likeT.pandas[0].predict(ent) print sign0(likeT.pandas[0].predict(ent)) # likeTS = ms.yaml2json('turtle_scripts/turtle_like.yml') # # print likeTS # likeT = ms.create_turtle(likeTS) # likeT.save() # ent = ents[0] # # print ents[0].generic() # ent._setattr('likeTravel', 'Y') # ms.crane.entityStore.save_all(ents) # print ent.generic()
def add_label(ent_id,field,value): ent = ms.load_entity(ent_id) ent._setattr(field,value) ms.crane.entityStore.save_one(ent)