When exporting to Excel, it is possible to add multiple number formats and then use set_column to apply any desired numeric format to any column. Attempting to do the same for date or date-time columns has no effect however.

It is possible to set a format for all date columns or all date-time columns using the date_format and datetime_format arguments to ExcelWriter(), but that applies a single format to all dates and a single format to all date-time values.

In my particular case, I have a column that is intended to represent a whole month, and I want to format it as such in the exported data, and I have other date columns that represent specific dates. There is apparently no way to accomplish that. Why?

Why not simply allow applying date formats to columns just as one can apply numeric formats to columns?

Comment From: stevecj

Suggested pattern:

Pass False as the value to date_format and/or datetime_format to indicate that the formats for columns of those types should not be set automatically.

Comment From: WillAyd

I think this would be a logical enhancement - interested in submitting a pull request @stevecj ?

Comment From: stevecj

I haven't worked on Pandas source before, but I'll certainly give it a shot.

Comment From: stevecj

Short on time. Still plan to try making this enhancement at some point if nobody beats me to it.

Comment From: shraddhazpy

If @stevecj isn't working on it, I would want to take it.

Comment From: MarcoGorelli

go ahead, thanks @shraddhazpy

Comment From: shraddhazpy

take

Comment From: shraddhazpy

Suggested pattern:

Pass False as the value to date_format and/or datetime_format to indicate that the formats for columns of those types should not be set automatically.

@MarcoGorelli just wanted to ensure that this is the pattern to be followed.

Comment From: MarcoGorelli

Looks like it - I think the suggestion is to make something analogous to float_format, but for dates