Excel and co. can parse a table from the clipboard, but pandas to_clip uses wrong format

[ghost]

No description provided.

[jreback]

what do you mean by wrong format?

it just writes text, does it need to have a sep for example?

[ghost]

It's been a while, iirc, just open up excel, copy a table and then paste it into an editor,
that's the format. pandas exports something else which excel can't parse.

Should be an easy fix.

[ghost]

Just remembered something else from way back, there's a way to get excel
to accept the data directly without going through the import wizard, I think it needs
to be UTF16LE+BOM. Probably requires some playing around to get it working.

Not necessary, but convenient (for excel anyway).

[jtratner]

Just needs to be tab delimited, instead of space delimited. (I was able to
do copy/paste directly with the current setup, it just puts it all in one
column for now):

col1\tcol2\tcol3\n
a\tb\tc\n

Really simple. I think current versions of Excel (i.e., 2007+) don't need
the header at all.

[ghost]

Should test with libreoffice as well. and unicode.

[jreback]

tab deliminted doesn't work when cross-exporting, between x and windows I made the defautl to do csv

[jtratner]

I think we settled on using tabs on Windows/mac and then commas in Linux?

[jtratner]

Someone care to test what works on Linux? Tabs on windows/mac means you can paste into excel and all the columns work well.

[ghost]

Tabs work fine on libreoffice in linux.

[jtratner]

@jreback that suggests '\t' is the correct choice. In what circumstance wouldn't you want that to be the default?