pySpark .join() with different column names and can't be hard coded before runtime

Question

I get this final = ta.join(tb, on=[\'ID\'], how=\'left\') both left an right have a \'ID\' column of the same name.

And I get this final = ta.join(