Export a LaTeX table from pandas DataFrame

Question

Is there an easy way to export a data frame (or even a part of it) to LaTeX?

I searched in google and was only able to find solutions using asciitables.

In addition to the replies here, this question has an interesting approach too. Export to csv and then import the csv into latex: http://tex.stackexchange.com/questions/54990/convert-numpy-array-into-tabular — Ricardo Magalhães Cruz, Nov 03 '15 at 14:17

score 107 · Accepted Answer · edited Oct 15 '21 at 14:46

107

DataFrames have a to_latex (see the pandas docs) method:

>>> df = pd.DataFrame(np.random.random((5, 5)))
>>> df  
          0         1         2         3         4
0  0.886864  0.518538  0.359964  0.167291  0.940414
1  0.834130  0.022920  0.265131  0.059002  0.530584
2  0.648019  0.953043  0.263551  0.595798  0.153969
3  0.207003  0.015721  0.931170  0.045044  0.432870
4  0.039886  0.898780  0.728195  0.112069  0.468485

>>> print(df.to_latex())
\begin{tabular}{|l|c|c|c|c|c|c|}
\hline
{} &         0 &         1 &         2 &         3 &         4 \\
\hline
0 &  0.886864 &  0.518538 &  0.359964 &  0.167291 &  0.940414 \\
1 &  0.834130 &  0.022920 &  0.265131 &  0.059002 &  0.530584 \\
2 &  0.648019 &  0.953043 &  0.263551 &  0.595798 &  0.153969 \\
3 &  0.207003 &  0.015721 &  0.931170 &  0.045044 &  0.432870 \\
4 &  0.039886 &  0.898780 &  0.728195 &  0.112069 &  0.468485 \\
\hline
\end{tabular}

You can simply write this to a tex file.

By default latex will render this as:

as it would appear in latex

Note: the to_latex (see the pandas docs) method offers several configuration options.

edited Oct 15 '21 at 14:46

iacob

20,084
6
92
119

answered Jan 17 '13 at 16:30

bmu

35,119
13
91
108

does `to_latex()` support multi-indexing? – TheChymera Nov 10 '13 at 17:31
1

@TheChymera: Yes. I just tried multi columns, which seems to work. Multi indices look a little bit strange to me. – bmu Nov 11 '13 at 18:46
8

The option `float_format=lambda x: '%10.2f' % x` proved to be very useful in my case. Also, in LaTeX, I had to add `\usepackage{booktabs}`, as the output uses `\toprule` which requires it. – metakermit Mar 18 '14 at 12:42
Great ! Does the reverse exists? I have a latex array and I would like to create a dataframe – Revolucion for Monica Aug 23 '18 at 17:25
Is there a way to end up with the pdf generated from the `latex` code? – Gathide Sep 17 '21 at 06:59

Thorsten Kranz · Answer 2 · 2013-01-17T13:58:43.053

10

Just write to a textfile. It's no magic:

import pandas as pd
df = pd.DataFrame({"a":range(10), "b":range(10,20)})
with open("my_table.tex", "w") as f:
    f.write("\\begin{tabular}{" + " | ".join(["c"] * len(df.columns)) + "}\n")
    for i, row in df.iterrows():
        f.write(" & ".join([str(x) for x in row.values]) + " \\\\\n")
    f.write("\\end{tabular}")

edited Jan 17 '13 at 13:58

answered Jan 17 '13 at 13:47

Thorsten Kranz

12,492
2
39
56

10

Although it is useful to show how easy it is to export tabular data to LaTeX using plain Python, it is of course much better to use existing facilities, i.e., `DataFrame.to_latex()` in this case. – hans_meine Mar 12 '14 at 10:49
6

Hi, I totally agree. My reasons were: 1. I started with pandas before version 0.9, where to_latex wasn't available 2. This might be useful when some adjustement is needed or someone wants to export to another format, e.g., XAML, custom XML or whatever. But thank you for your comment – Thorsten Kranz Mar 12 '14 at 12:55

score 10 · Answer 3 · answered Aug 06 '18 at 16:03

10

If you want to save it:

with open('mytable.tex', 'w') as tf:
     tf.write(df.to_latex())

answered Aug 06 '18 at 16:03

Armin Alibasic

261
4
9

19

Actually, you just need `df.to_latex('mytable.tex')` – Diogo Oct 06 '19 at 06:31

Export a LaTeX table from pandas DataFrame

3 Answers3

Linked