Aligning individual columns in pandas to_latex

Question

I'm using the pandas to_latex method to convert a dataframe into a latex tabular. I don't see an option to change the alignment fields of the resulting tabular. For example, I have a dataframe that looks like this:

In [46]: df
Out[46]: 
    Number of days  Tuples Distinct Tuples
162             29  700587           41300
163             20  497599           29302
164             15  365599           21382
165             10  256903           14916
166              5  127647            7441
167              2   54254            3117
168              1   26987            1288

and my output table looks like this:

In [50]: print df.to_latex(index=None)
\begin{tabular}{lll}
\toprule
Number of days &  Tuples & Distinct Tuples \\
\midrule
            29 &  700587 &           41300 \\
            20 &  497599 &           29302 \\
            15 &  365599 &           21382 \\
            10 &  256903 &           14916 \\
             5 &  127647 &            7441 \\
             2 &   54254 &            3117 \\
             1 &   26987 &            1288 \\
\bottomrule
\end{tabular}

What I want is for the {lll} alignment to change to {rrr}. In general, I might even want different alignments for different columns, or even use a vertical separator | in a {r|r|r} designator.

Is this currently supported ?

Answer 1

By now (pandas version 0.17.1), the to_latex method has the column_format parameter, so you could simply do

print df.to_latex(index=None, column_format='rrr')

Answer 2

As you can see in code of pandas(this function is used for rendering latex_table): https://github.com/pydata/pandas/blob/master/pandas/core/format.py#L492 There's not supported yet. Only if your data is numpy's number it will be formatted to right. In python it's pretty easy to format your columns.

print df.to_latex(index=None).replace('lll','rrr')

or in more generic way using regex replace.