How to explicitly set database engine when using to_sql() in pandas

Question

How could I possibly patch the sql statement in the pandas to_sql() function, so the newly created table uses the MYISAM storage engine?

I need MYISAM because of a very large amount of columns. This currently causes issues with the standard database engine INNODB (Row size too large (> 8126).

I am aware that it is possible to set the database engine explicitly in a mysql statement upon table creation. Perhaps it's possible to patch the sql that is generated by the to_sql() function?

To specify explicitly that you want a MyISAM table, indicate that with an ENGINE table option: CREATE TABLE t (i INT) ENGINE = MYISAM;

This is how I create the table currently

df.to_sql(con=engine, name="generated_" + reportConfiguration.shortName + "_" + reportConfiguration.marketplace, if_exists='replace',index=False)

Answer 1

As per my knowledge, storage engine of MySQL can't be set using df.to_sql or engine = create_engine('mysql+pymy....://x@y/z') .

MySQL storage engine can be added to table structure once it is created.
executing an alter table command over the engine connection, you can modify the table storage engine.

Example :

with engine.begin() as conn:     
      conn.execute("ALTER TABLE table_name ENGINE = MYISAM")

Documentation :

• Working with Engines and Connections
  • execute(object, *multiparams, **params)
  • Executes a SQL statement construct and returns a ResultProxy