update

SmartDataAnalytics · Jul 8, 2019 · cdf408d · cdf408d
1 parent fd88b1e
commit cdf408d
Show file tree

Hide file tree

Showing 3 changed files with 1 addition and 6 deletions.
diff --git a/README.md b/README.md
@@ -39,6 +39,7 @@ python -m spacy download en
 python -m spacy download en_core_web_lg
 
 python preprocess_kb_2.py --data_dir soccer_conversations/
+python preprocess_kb_incar.py --data_dir preproc_files/incar/ --stoi vocab/w2i_incar.npy --vocab_glove vocab/glove300_incar.npy
 python utils/generate_entities_soccer.py
 ```
 

diff --git a/kg_build/build_KG_clubs.py b/kg_build/build_KG_clubs.py
@@ -30,7 +30,6 @@
 }
 
 
-
 IGONRE_PLAYER_STATE = ["INJ","OTH","SUS","PRE","RET"]
 
 #players.append(['Number', 'Position', 'Name', 'Date of birth', 'height', 'senior career caps', 'senior career goals'])
@@ -176,13 +175,10 @@ def side_bar_info(wiki_link):
         if tr.find_all('th'):
             if tr.find_all('th')[0].get_text().strip()=="Manager":
                 head_coach = tr.find_all('td')[0].get_text().strip()
-
             if tr.find_all('th')[0].get_text().strip()=="Head coach":
                 head_coach = tr.find_all('td')[0].get_text().strip()
-
             if tr.find_all('th')[0].get_text().strip()=="President":
                 president = tr.find_all('td')[0].get_text().strip()
-
             if tr.find_all('th')[0].get_text().strip()=="Chairman":
                 president = tr.find_all('td')[0].get_text().strip()
 
@@ -250,8 +246,6 @@ def fetch_current_squad(club):
 
 
 
-
-
 if __name__ == "__main__":
 
     base_url = 'https://en.wikipedia.org/wiki/'

diff --git a/preproc_files/incar/valid.npy b/preproc_files/incar/valid.npy