How to show a table with both head and tail?

[Norlandz]

ie:
Is there a config such that:
instead of df.head(5).append(df.tail(5)),
just df, and both head and tail are shown (with ... in middle)?

[mwouts]

Hello @Norlandz , that's a very interesting question, and I would have use of that myself.

It is currently not possible to do so. However seeing https://datatables.net/forums/discussion/60989/fixed-first-row, I have experimented a bit with table.header and table.footer and was able to get this, with one fixed row on the top, one fixed row at the bottom, and the whole the table in between (which you can scroll):

Would that work for you? (with maybe a few more fixed rows at the top/bottom, and no row duplication)

For reference the code for that experiment is available at https://live.datatables.net/qujunigu/856/edit

[Norlandz]

If this is the current workaround, I may try with it.

But if you think this is not worth the time to implement it, its fine for me too.

--
Can it display 5 tail rows in footer?

--
Is there an api for writing those js in ipynb?

--
To solve this, wouldnt there be a way that:
you can sort the index in a way that
shows 1,2,3,97,98,99
instead of 1,2,3,4,5,6

(--ie: in the code, for the comparator, give a higher priority for the last indexes 97,98,99)

--
(Thanks for your reply!)

[mwouts]

Oh it's not even a workaround - just a preview of how a future feature could look like.

Sure it would make sense to be able to fix multiple rows both in the header and footer. Would two options like fixed_first_rows and fixed_last_rows, or even just fixed_rows (type int, default to 0) fit your use case ? I don't think I would like to support arbitrary fixed rows indices, since the fixed rows will be always displayed at the top or bottom of the datatables, and I prefer to preserve the ordering of the input dataframe.

[Norlandz]

@mwouts

Oh it's not even a workaround

(To me, it feels a workaround.)

fixed_first_rows and fixed_last_rows

Sounds good for me.

(Thanks for the reply!)