In [1]:

    

import pandas as pd
pd.set_option('display.max_rows', 100)


    

In [2]:

    

from horacy import HoracyModel
model = HoracyModel('model_4000/').load()


    

    




load done in 2.69 seconds

In [25]:

    

results = model.tfidf_query('ventil')
df = pd.DataFrame(zip(*results),columns=['id','token','df'])
df


    

    
Out[25]:







  

    

      

      
id
      
token
      
df
    
  
  

    

      
0
      
3719
      
ventilation
      
804
    
    

      
1
      
23641
      
ventilated
      
155
    
    

      
2
      
9440
      
ventilator
      
94
    
    

      
3
      
29220
      
ventilators
      
85
    
    

      
4
      
18845
      
ventilatory
      
54
    
    

      
5
      
52285
      
ventilator-associated
      
50
    
    

      
6
      
79235
      
ventilation:
      
11
    
    

      
7
      
23642
      
ventilator-induced
      
10
    
    

      
8
      
87766
      
hypoventilation
      
8
    
    

      
9
      
67896
      
non-ventilated
      
7
    
    

      
10
      
37127
      
ventilator-free
      
6
    
    

      
11
      
46658
      
ventilator-days
      
6
    
    

      
12
      
77019
      
ventilation-induced
      
5
    
    

      
13
      
93910
      
hyperventilation
      
5
    
    

      
14
      
23006
      
well-ventilated
      
4
    
    

      
15
      
47508
      
ventilatorassociated
      
4
    
    

      
16
      
52963
      
ventilating
      
4
    
    

      
17
      
63628
      
ventilatorinduced
      
4
    
    

      
18
      
78097
      
nurse:ventilated
      
3
    
    

      
19
      
85030
      
g006ventilation
      
3
    
  

In [30]:

    

results = model.tfidf_query('mechanical')
df = pd.DataFrame(zip(*results),columns=['id','token','df'])
df


    

    
Out[30]:







  

    

      

      
id
      
token
      
df
    
  
  

    

      
0
      
3701
      
mechanical
      
804
    
    

      
1
      
10943
      
mechanically
      
150
    
    

      
2
      
71211
      
biomechanical
      
9
    
    

      
3
      
84684
      
optomechanical
      
9
    
    

      
4
      
53365
      
electromechanical
      
4
    
    

      
5
      
84734
      
nanomechanical
      
4
    
  

In [8]:

    

def multi_query(*args):
    ids = set()
    for i,arg in enumerate(args):
        query,_,exclude = arg.partition(' ~~ ')
        q_ids = model.inverted_query(query,exclude)
        if i==0:
            ids.update(q_ids)
        else:
            ids.intersection_update(q_ids)
    return ids

def show_results(ids):
    print(f"RESULTS: {len(ids)}\n\n")
    for id in ids:
        doc = model.get_doc(id)
        text = model.doc_to_text(doc).replace('\n',' ').replace('\r',' ')
        print(f"id:{id}\n{text}\n\n")

import re
def filter_results(ids,query):
    q = re.compile(query)
    for id in ids:
        doc = model.get_doc(id)
        text = model.doc_to_text(doc).replace('\n',' ').replace('\r',' ')
        for sen in model.text_to_sentences(text):
            if q.findall(sen):
                print(id,sen)
        
ids = multi_query('mechanical','ventil')
filter_results(ids,'\b(age|old|year)')


    

In [ ]: