pandas dataframe to sql - append data to exisitng table without duplication

Question

I have a MySQL Table feinstaub with the Columns (created_at,PM 2.5,PM 10,entry_id) c4 is unique. And i have a pandas dataFrame with equal column names. In this dataframe are new values and already existing values compare to the sql table. I use this line to send the dataframe to sql server.

df.to_sql("Feinstaub", con=engine, if_exists="append", index=False)

It worked only when in the dataframe are no duplicated vlaues. If there are some dupl. values it doesent work. I found this solution: Pandas to_sql() to update unique values in DB?

with engine.begin() as cn:
   sql = """INSERT INTO myFinalTable (Col1, Col2, Col3, ...)
            SELECT t.Col1, t.Col2, t.Col3, ...
            FROM myTempTable t
            WHERE NOT EXISTS 
                (SELECT 1 FROM myFinalTable f
                 WHERE t.MatchColumn1 = f.MatchColumn1
                 AND t.MatchColumn2 = f.MatchColumn2)"""

   cn.execute(sql)

I end up with this one:

df.to_sql("temp_feinstaub_wohnzimmer", con=engine, if_exists="replace", index=False)
with engine.begin() as cn:
   sql = """INSERT INTO feinstaub (created_at, 'PM 2.5' , 'PM 10', entry_id)
            SELECT t.Column1, t.Column2, t.Column3 ,t.Column4
            FROM temp_feinstaub_wohnzimmer t
            WHERE NOT EXISTS
                (SELECT 1 FROM feinstaub f
                 WHERE t.MatchColumn1 = f.MatchColumn1
                 AND t.MatchColumn2 = f.MatchColumn2
                 AND t.MatchColumn3 = f.MatchColumn3
                 AND t.MatchColumn4 = f.MatchColumn4)"""

   cn.execute(sql)

It raised an sql syntax error. I try to rename f.MatchColumn also but still give me an sql syntax error?

EDIT: I use this code now and it worked with the backticks thanks! But it raised another error ;)

#Send the Data to SQL database
df.to_sql("temp_feinstaub_wohnzimmer", con=engine, if_exists="replace", index=False)
with engine.begin() as cn:
   sql = """INSERT INTO feinstaub (created_at, `PM 2.5` , `PM 10`, entry_id)
            SELECT t.created_at, t.`PM 2.5`, t.`PM 10` ,t.entry_id
            FROM temp_feinstaub_wohnzimmer t
            WHERE NOT EXISTS
                (SELECT 1 FROM feinstaub f
                 WHERE t.created_at = f.created_at
                 AND t.`PM 2.5` = f.`PM 2.5`
                 AND t.`PM 10` = f.`PM 10`
                 AND t.entry_id = f.entry_id)"""

   cn.execute(sql)

and now i get the folloing error:

sqlalchemy.exc.IntegrityError: (_mysql_exceptions.IntegrityError) (1062, "Duplicate entry '3825' for key 'entry_id'") [SQL: 'INSERT INTO feinstaub_wohnzimmer (created_at, `PM 2.5` , `PM 10`, entry_id)\n            SELECT t.created_at, t.`PM 2.5`, t.`PM 10` ,t.entry_id\n            FROM temp_feinstaub_wohnzimmer t\n            WHERE NOT EXISTS\n                (SELECT 1 FROM feinstaub_wohnzimmer f\n                 WHERE t.created_at = f.created_at\n                 AND t.`PM 2.5` = f.`PM 2.5`\n                 AND t.`PM 10` = f.`PM 10`\n                 AND t.entry_id = f.entry_id)']

And what is the syntax error? Column names should not be single quoted. Use backticks to escape spaces and special characters. Also, obviously, you have to rename *MatchColumn* to actual column names which are placeholders here. Add or remove if more or less than 4 columns. — Parfait, Dec 06 '18 at 22:00

score 3 · Answer 1 · answered Dec 07 '18 at 11:26

With this it works for me... I can excecute the script several times and only the new values find their way into the mysql database.

from sqlalchemy import exc
num_rows = len(df)
#Iterate one row at a time
for i in range(num_rows):
    try:
        #Try inserting the row
        df.iloc[i:i+1].to_sql(name="feinstaub_wohnzimmer",con = engine,if_exists = 'append',index=False)
    except exc.IntegrityError:
        #Ignore duplicates
        pass

pandas dataframe to sql - append data to exisitng table without duplication

1 Answers1