DataFrame.to_csv sometimes ignores header argument

[marcua]

Using PANDAS 0.13.1.

If you call DataFrame.to_csv with index=False and a header, the header doesn't replace the column names. With index=True, header correctly replaces the csv header.

The code behind this: https://github.com/pydata/pandas/blob/master/pandas/core/format.py#L1080

Presumably we want to replace cols with headers even in the else on line 1091.

Happy to take a stab at fixing this, but didn't know if I'm missing some interaction between the variables that makes this not a bug. Thanks!

[jreback]

pretty sure that their is an issue about this
care to find and link the issue?

[marcua]

Looks like it's this: #6186

Unsure how to mark this as a duplicate, but happy to tackle the solution if you agree it's a problem :)

[TomAugspurger]

We can just close this one now that you've linked to the original.