# setup: import pandas as pd ; import numpy ; N = 1000000 ; df1 = pd.DataFrame(numpy.random.randint(0, 10, size=(N, 3)), columns=["col1", "col2", "col3"]) ; df2 = pd.DataFrame(numpy.random.randint(0, 10, size=(N, 3)), columns=["col4", "col5", "col6"]) # noqa: E501 # run: join_df(df1, df2) def join_df(df1, df2): return df1.join(df2, how="outer")