handling spaCy entities

spaCy-entities.py

import spacy
from spacy import displacy

path = "en_core_web_sm"

nlp = spacy.load(path)

path_folder = "/path/to/file/"

import pandas as pd

df = pd.read_csv(path_folder + "file.csv")

# convert entities to list

docs = list()
for index, row in df.iterrows():
    docs.append(row['text'], dict(id = row['id']))
    
%%time
r = []
doc = list()

for doc, context in nlp.pipe(docs, as_tuples=True):
    add = [(ent.text, ent.label_, ent.start_char, ent.end_char, context['id'])
    r.append(add)
    
# run to csv
l = []
for i in r:
    p = pd.DataFrame(i, columns=['text','label','start_char','end_char','id'])
    l.append(p)

dfr = pd.concat(l, ignore_index = True)
dfr.to_csv(path_folder + "entities.csv")