Essence, the 12 major Pandas common configuration skills, you may not understand.

Hello, in the process of using Pandas, in addition to data, we deal more with tables. In order to better display a table data, there must be good settings in the early stage.

This article introduces the common configuration skills of Pandas, mainly based on options and settings. Welcome to collect and learn, like praise and support.

Note: technical exchange group is provided at the end of the text


This is a conventional import method!

import pandas as pd

Ignore warning

Due to the update of the version, some usage of Pandas may be removed in the near future. Some warnings (not errors) often appear. The relevant warnings can be ignored with the following code:

# Ignore warning
import warnings

float data accuracy

View default precision

The default is to keep 6 decimal places. Print the current accuracy by:

pd.get_option( 'display.precision')

Modify accuracy

Set the precision to 2 bits

pd.set_option( 'display.precision',2)
# Writing 2: PD options. display. precision = 2

Then we print again, and the current accuracy becomes 2 bits:

pd.get_option( 'display.precision')

Display rows

The default number of rows displayed is 60

pd.get_option("display.max_rows")  # The default is 60

The default minimum number of rows is 10 bits:

pd.get_option("display.min_rows")  # Minimum display rows

Modify the number of rows displayed

Modify the maximum number of displayed lines to 999, and then view:

pd.set_option("display.max_rows",999)  # Maximum number of rows to display

Modify the minimum number of rows displayed:


Reset function

Reset using reset_ After the option method, the setting will change to the default form (numerical value):

pd.get_option("display.max_rows")  # Back to 60
pd.get_option("display.min_rows")  # Back to 10

Regular function

If we modify the settings of multiple options and want to restore them at the same time, we can reset multiple options using regular expressions.

Here, it means that all settings starting with display are reset:

# ^It means to start with a character. Here, it means to start all resets with display

reset all

If all is used, all settings are reset:


Display column

Since you can control the number of rows displayed, you can also control the number of columns displayed

View the number of columns displayed

The number of columns displayed by default is 20:


#Another way to write: through attributes

Change the number of columns

Modify the number of columns displayed to 100:

# Change to 100

To view the number of modified columns:

# View modified values

Show all columns

If set to None, all columns are displayed:




Modify column width

The above is to view the number of columns, and the following is to set the width of each column. The width of a single column of data, calculated in the number of characters, is indicated by an ellipsis when it exceeds.

Default column width

The default column width is 50 characters:

pd.get_option ('display.max_colwidth')

Modify column width

Modify the displayed column width to 100:

# Change to 100
pd.set_option ('display.max_colwidth', 100)

To view the displayed column width and length:

pd.get_option ('display.max_colwidth')

Show all columns

Show all columns:

pd.set_option ('display.max_colwidth', None)

Folding function

Whether to collapse when the output data width exceeds the set width. Generally, False is not used to collapse, but True is used to collapse.

pd.set_option("expand_frame_repr", True)  # fold
pd.set_option("expand_frame_repr", False)  # No folding

Snippet modification settings

The settings described above, if modified, are the of the whole environment; We can also make temporary settings for only one code block.

If you run out of the current code block, it will fail and return to the original setting.

Suppose this is the first code block:


Here is the second code block:

# Set the current code block

with pd.option_context("display.max_rows", 20, "display.max_columns", 10):

Here is the third code block:


In the above example, we can find that the setting is invalid outside the specified code block

Number formatting

There is a display in Pandas float_ Format method can format and output floating-point numbers, such as expressed in thousandths, percentages, fixed decimal places, etc.

You can also use this method if other data types can be converted to floating point numbers.

The callable should accept a floating point number and return a string with the desired format of the number

Millennial representation

When the data is relatively large, it is hoped to express the data in the form of thousandths, which is clear at a glance:

df = pd.DataFrame({
    "percent":[12.98, 6.13, 7.4],


Special symbols

In addition to the% sign, we can also use other special symbols:

Zero threshold conversion

What does threshold conversion mean? First, the implementation of this function uses display chop_ Threshold method.

Indicates the threshold for displaying data in Series or DF as a number. If it is greater than this number, it will be displayed directly; If less than, it is displayed with 0.

Change drawing method

By default, pandas uses matplotlib as the drawing back end. We can modify the settings:

import matplotlib.pyplot as plt
%matplotlib inline

# Default
df1 = pd.DataFrame(dict(a=[5,3,2], b=[3,4,1]))

Change the back end of the drawing to become a powerful plot:

# Writing method 1
pd.options.plotting.backend = "plotly"

df = pd.DataFrame(dict(a=[5,3,2], b=[3,4,1]))
fig = df.plot()

# Writing method 2
df = pd.DataFrame(dict(a=[5,3,2], b=[3,4,1]))
fig = df.plot(backend='plotly') # Specify here

Modify column head alignment direction

By default, the attribute fields (column headers) are aligned to the right and can be set. Here is an example from the official website:

Print out the current settings and reset all options

pd.describe_option() prints all current settings and resets all options. Here are some setting options:

Configuration skills

Common configurations are summarized below, which can be used after copying:

import pandas as pd  # International practice

import warnings
warnings.filterwarnings('ignore')  # Ignore warnings in text

pd.set_option( 'display.precision',2)
pd.set_option("display.max_rows",999)  # Maximum number of rows to display
pd.set_option("display.min_rows",20)   # Minimum number of rows to display
pd.set_option('display.max_columns',None)  # All columns
pd.set_option ('display.max_colwidth', 100)   # Modify column width
pd.set_option("expand_frame_repr", True)  # fold
pd.set_option('display.float_format',  '{:,.2f}'.format)  # Thousandth
pd.set_option('display.float_format', '{:.2f}%'.format)  # Percentage form
pd.set_option('display.float_format', '{:.2f}¥'.format)  # Special symbols
pd.options.plotting.backend = "plotly"  # Modify drawing
pd.set_option("colheader_justify","left")  # Column field alignment
pd.reset_option('all')  # Reset

Technical exchange

Welcome to reprint, collect, gain, praise and support!

At present, a technical exchange group has been opened, with more than 2000 group friends. The best way to add notes is: source + Interest direction, which is convenient to find like-minded friends

  • Method ① send the following pictures to wechat, long press identification, and the background replies: add group;
  • Mode ②. Add micro signal: dkl88191, remarks: from CSDN
  • WeChat search official account: Python learning and data mining, background reply: add group

Keywords: Python Data Analysis Data Mining

Added by desenhistas on Sat, 18 Dec 2021 08:22:00 +0200