Add an ends function that shows both head and tail of the df

[icfly2]

Add the following function to pd. dataFrame and pd.Series

def ends(df, x=5):
    """Returns both head and tail of the dataframe or series.
    
    Args:
        x (int): Optional number of rows to return for each head and tail
    """
    print('{} rows x {} columns'.format(np.shape(df)[0],np.shape(df)[1]))
    return df.head(x).append(df.tail(x))

Problem description

Often both the beginning and end of a df are of interest, fore example in a time series.

This leads to calling df.head() df.tail() in two seperate notebook cells. This is not only tedious, but also leads to a cluttered notebook. A function that returns both of these + a print on the number of rows and columns, thus allowing a check if the index matches the number of rows.

Example

import pandas as pd
import numpy as np
df = pd.DataFrame(np.random.rand(1500,6))
print(ends(df,2))

1500 rows x 6 columns

	0	1	2	3	4	5
0	0.949695	0.160928	0.434134	0.943103	0.477830	0.903479
1	0.736711	0.103746	0.028694	0.205910	0.226061	0.458452
1498	0.362950	0.586887	0.399681	0.115366	0.239049	0.386281
1499	0.018102	0.852198	0.880993	0.671604	0.705586	0.802237

Output of pd.show_versions()

INSTALLED VERSIONS

commit: None
python: 3.6.2.final.0
python-bits: 64
OS: Windows
OS-release: 8.1
machine: AMD64
processor: Intel64 Family 6 Model 61 Stepping 4, GenuineIntel
byteorder: little
LC_ALL: None
LANG: None
LOCALE: None.None

pandas: 0.21.0
pytest: 3.3.1
pip: 9.0.1
setuptools: 38.2.4
Cython: 0.26.1
numpy: 1.12.1
scipy: 0.19.1
pyarrow: None
xarray: None
IPython: 6.1.0
sphinx: 1.6.3
patsy: 0.4.1
dateutil: 2.6.1
pytz: 2017.3
blosc: None
bottleneck: 1.2.1
tables: 3.4.2
numexpr: 2.6.2
feather: None
matplotlib: 2.0.2
openpyxl: 2.4.8
xlrd: 1.1.0
xlwt: 1.3.0
xlsxwriter: 0.9.8
lxml: 3.8.0
bs4: 4.6.0
html5lib: 0.999999999
sqlalchemy: 1.1.13
pymysql: None
psycopg2: None
jinja2: 2.10
s3fs: None
fastparquet: None
pandas_gbq: None
pandas_datareader: None

[jorisvandenbossche]

I thought we discussed this before, but can't find such issue.

That said, this is closely related to #17023. Because if we reduce the default number of rows, I think such a feature becomes less needed IMO.

In general, we are not very keen on adding more methods to DataFrame, unless there is a very clear need.

[TomAugspurger]

FWIW, my utils file has a context manager for this:

def show(df):
    with pd.option_context("display.max_rows", 10):
        display(df)

And I called it with df.pipe(show). Agreed that we shouldn't add a new method, especially given #17023.

[jreback]

see #4193 and #1889

[jreback]

this seems quite duplicate of what we already have, by default the repr already does this. using display.max_rows to control the break of this.

[icfly2]

Perhaps it might be worth adding a line to the docs as it seems quite an obscure but very useful trick.

I would suggest referring to head in the docs for tail and vice versa and adding the comment about display.max_rows

I'm happy to try my hand at that addition if there is some general support for the idea.

[jreback]

@icfly2 #18749 already is updating he doc-strings, so natural to add there.

Love to have you contribute on other issues!

[adamrossnelson]

Worth a mention that this issue also relates to: #42837

The idea is that there is a way to preserve vertical screen space when inspecting 'tall' (many rows, few cols) data frames.