BUG: Setting frame into df with dup cols loses dtypes by phofl · Pull Request #53143 · pandas-dev/pandas (original) (raw)
df = DataFrame([[1, 2, 3, 4], [4, 5, 6, 7]], columns=["a", "b", "a", "c"])
df = DataFrame([[1, 2, 3, 4], [4, 5, 6, 7]], columns=["a", "b", "a", "c"])