漂亮打印熊猫数据框

How can I print a pandas dataframe as a nice text-based table, like the following?

+------------+---------+-------------+
| column_one | col_two |   column_3  |
+------------+---------+-------------+
|          0 |  0.0001 | ABCD        |
|          1 |  1e-005 | ABCD        |
|          2 |  1e-006 | long string |
|          3 |  1e-007 | ABCD        |
+------------+---------+-------------+

I’ve just found a great tool for that need, it is called tabulate.

It prints tabular data and works with DataFrame.

from tabulate import tabulate
import pandas as pd

df = pd.DataFrame({'col_two' : [0.0001, 1e-005 , 1e-006, 1e-007],
                   'column_3' : ['ABCD', 'ABCD', 'long string', 'ABCD']})
print(tabulate(df, headers='keys', tablefmt='psql'))

+----+-----------+-------------+
|    |   col_two | column_3    |
|----+-----------+-------------|
|  0 |    0.0001 | ABCD        |
|  1 |    1e-05  | ABCD        |
|  2 |    1e-06  | long string |
|  3 |    1e-07  | ABCD        |
+----+-----------+-------------+

Note:

To suppress row indices for all types of data, pass showindex="never" or showindex=False.

A simple approach is to output as html, which pandas does out of the box:

df.to_html('temp.html')

pandas >= 1.0

If you want an inbuilt function to dump your data into some github markdown, you now have one. Take a look at to_markdown:

df = pd.DataFrame({"A": [1, 2, 3], "B": [1, 2, 3]}, index=['a', 'a', 'b'])  
print(df.to_markdown()) 

|    |   A |   B |
|:---|----:|----:|
| a  |   1 |   1 |
| a  |   2 |   2 |
| b  |   3 |   3 |

Here’s what that looks like on github:

Note that you will still need to have the tabulate package installed.

If you are in Jupyter notebook, you could run the following code to interactively display the dataframe in a well formatted table.

This answer builds on the to_html(‘temp.html’) answer above, but instead of creating a file displays the well formatted table directly in the notebook:

from IPython.display import display, HTML

display(HTML(df.to_html()))

Credit for this code due to example at: Show DataFrame as table in iPython Notebook

You can use prettytable to render the table as text. The trick is to convert the data_frame to an in-memory csv file and have prettytable read it. Here’s the code:

from StringIO import StringIO
import prettytable    

output = StringIO()
data_frame.to_csv(output)
output.seek(0)
pt = prettytable.from_csv(output)
print pt

I used Ofer’s answer for a while and found it great in most cases. Unfortunately, due to inconsistencies between pandas’s to_csv and prettytable‘s from_csv, I had to use prettytable in a different way.

One failure case is a dataframe containing commas:

pd.DataFrame({'A': [1, 2], 'B': ['a,', 'b']})

Prettytable raises an error of the form:

Error: Could not determine delimiter

The following function handles this case:

def format_for_print(df):    
    table = PrettyTable([''] + list(df.columns))
    for row in df.itertuples():
        table.add_row(row)
    return str(table)

If you don’t care about the index, use:

def format_for_print2(df):    
    table = PrettyTable(list(df.columns))
    for row in df.itertuples():
        table.add_row(row[1:])
    return str(table)

Following up on Mark’s answer, if you’re not using Jupyter for some reason, e.g. you want to do some quick testing on the console, you can use the DataFrame.to_string method, which works from — at least — Pandas 0.12 (2014) onwards.

import pandas as pd

matrix = [(1, 23, 45), (789, 1, 23), (45, 678, 90)]
df = pd.DataFrame(matrix, columns=list('abc'))
print(df.to_string())

#  outputs:
#       a    b   c
#  0    1   23  45
#  1  789    1  23
#  2   45  678  90

Maybe you’re looking for something like this:

def tableize(df):
    if not isinstance(df, pd.DataFrame):
        return
    df_columns = df.columns.tolist() 
    max_len_in_lst = lambda lst: len(sorted(lst, reverse=True, key=len)[0])
    align_center = lambda st, sz: "{0}{1}{0}".format(" "*(1+(sz-len(st))//2), st)[:sz] if len(st) < sz else st
    align_right = lambda st, sz: "{0}{1} ".format(" "*(sz-len(st)-1), st) if len(st) < sz else st
    max_col_len = max_len_in_lst(df_columns)
    max_val_len_for_col = dict([(col, max_len_in_lst(df.iloc[:,idx].astype('str'))) for idx, col in enumerate(df_columns)])
    col_sizes = dict([(col, 2 + max(max_val_len_for_col.get(col, 0), max_col_len)) for col in df_columns])
    build_hline = lambda row: '+'.join(['-' * col_sizes[col] for col in row]).join(['+', '+'])
    build_data = lambda row, align: "|".join([align(str(val), col_sizes[df_columns[idx]]) for idx, val in enumerate(row)]).join(['|', '|'])
    hline = build_hline(df_columns)
    out = [hline, build_data(df_columns, align_center), hline]
    for _, row in df.iterrows():
        out.append(build_data(row.tolist(), align_right))
    out.append(hline)
    return "\n".join(out)


df = pd.DataFrame([[1, 2, 3], [11111, 22, 333]], columns=['a', 'b', 'c'])
print tableize(df)

Output:
+-------+----+-----+
|    a  |  b |   c |
+-------+----+-----+
|     1 |  2 |   3 |
| 11111 | 22 | 333 |
+-------+----+-----+

I wanted a paper printout of a dataframe but I wanted to add some results and comments as well on the same page. I have worked through the above and I could not get what I wanted. I ended up using file.write(df1.to_csv()) and file.write(“,,,blah,,,,,,blah”) statements to get my extras on the page. When I opened the csv file it went straight to a spreadsheet which printed everything in the right pace and format.