In my embedding table created with PGVector I added data from another table so that I could perform the varied searches correctly. However, I forgot to add one of the columns from the other table, the “language” column. Is it possible for me to just add this column without overwriting what already exists?

Next, I will show the code I used to perform the operation.

import psycopg2
from langchain_core.documents import Document

try:
     conn = psycopg2.connect(connection_string)
     
     cursor = conn.cursor()
    
     cursor.execute("""
     SELECT id_company_training, company_id, dialog, question, answer, "language" FROM company_training WHERE type = 'COMPANY_INTENTS'
     """)

     records = cursor.fetchall()
     
     for row in records:
        id_company_training, company_id, dialog, question, answer = row
        if len(question)>0:
            vector_db.add_documents([Document(page_content=question, metadata={
            'id_company_training':id_company_training,
            'company_id': company_id,
            'dialog': dialog,
            'question': question,
            'answer': answer
            })]) 
     
     cursor.close()
     conn.close()
 
except Exception as e:
     print(f"Erro ao conectar ao banco de dados: {e}")

I haven’t performed any execution yet because I’m afraid of overwriting my table.