ENH: to_latex(): add posibility to define the latex environment (instead of table)

[horror-vacui]

Is your feature request related to a problem?

I wish pandas's to_latex() function could create table* environment instead of table for wide tables. The generated table sometimes does not fit into one column of a two column document.

Describe the solution you'd like

A parameter where one can define the name of the environment. E.g. df.to_latex(env="foobar*") to generate: \begin{foobar*} table stuff here \end{foobar*}.

API breaking implications

One more parameter.

Describe alternatives you've considered

Write a code, which replaces begin{table} and end{table} to the desired environment. (table* in most of the cases)

[bashtage]

Why not just df.to_latex().replace("{table}","{table*}")? Since you aren't proposing fundamentally changing the rendering of the table, it seems like there is not a strong case for expanding the API.

[horror-vacui]

Why not just df.to_latex().replace("{table}","{table*}")? Since you aren't proposing fundamentally changing the rendering of the table, it seems like there is not a strong case for expanding the API.

I would be happy to use something like that, but the snippet you wrote, does not work with the latest release (1.1.3).

Traceback (most recent call last):
  File "soa.py", line 236, in <module>
    column_format="cccccccccccccc"
AttributeError: 'NoneType' object has no attribute 'replace'

for

df_table.to_latex(
            buf=tex_out,
            index=False,
            caption="Comparison with the State of the Art",
            label="tab:soa",
            na_rep="---",
            escape=False,
            bold_rows=True,
            column_format="cccccccccccccc"
            ).replace("{table}","{table*}")

[bashtage]

You need to not pass buf to get a string, which you can then call replace on.

buf.write(
    df_table.to_latex(
        index=False,
        caption="Comparison with the State of the Art",
        label="tab:soa",
        na_rep="---",
        escape=False,
        bold_rows=True,
        column_format="cccccccccccccc",
    ).replace("{table}", "{table*}")
)

[Kreijstal]

I would like to have this functionality, you don't necessarily need a tabular envirionment, it could be a matrix, a pmatrix, or some other environment. Or maybe you don't want any headers at all!

[Kreijstal]

@jreback there is still no option to allow no environment at all, there are only 2 environments allowed?