选项和设置#

概述#

pandas 提供了一个选项 API,用于配置和自定义与 DataFrame 显示、数据行为等相关的全局行为。

选项具有完整的“点分隔式”不区分大小写的名称(例如 display.max_rows)。您可以直接作为顶级 options 属性的属性获取/设置选项。

In [1]: import pandas as pd

In [2]: pd.options.display.max_rows
Out[2]: 15

In [3]: pd.options.display.max_rows = 999

In [4]: pd.options.display.max_rows
Out[4]: 999

API 由 5 个相关函数组成,可直接从 pandas 命名空间访问。

注意

开发者可以查看 pandas/core/config_init.py 获取更多信息。

以上所有函数都接受一个正则表达式模式 (re.search 风格) 作为参数,以匹配一个明确的子字符串。

In [5]: pd.get_option("display.chop_threshold")

In [6]: pd.set_option("display.chop_threshold", 2)

In [7]: pd.get_option("display.chop_threshold")
Out[7]: 2

In [8]: pd.set_option("chop", 4)

In [9]: pd.get_option("display.chop_threshold")
Out[9]: 4

以下将不起作用,因为它匹配多个选项名称,例如 display.max_colwidthdisplay.max_rowsdisplay.max_columns

In [10]: pd.get_option("max")
---------------------------------------------------------------------------
OptionError                               Traceback (most recent call last)
Cell In[10], line 1
----> 1 pd.get_option("max")

File ~/work/pandas/pandas/pandas/_config/config.py:274, in CallableDynamicDoc.__call__(self, *args, **kwds)
    273 def __call__(self, *args, **kwds) -> T:
--> 274     return self.__func__(*args, **kwds)

File ~/work/pandas/pandas/pandas/_config/config.py:146, in _get_option(pat, silent)
    145 def _get_option(pat: str, silent: bool = False) -> Any:
--> 146     key = _get_single_key(pat, silent)
    148     # walk the nested dict
    149     root, k = _get_root(key)

File ~/work/pandas/pandas/pandas/_config/config.py:134, in _get_single_key(pat, silent)
    132     raise OptionError(f"No such keys(s): {repr(pat)}")
    133 if len(keys) > 1:
--> 134     raise OptionError("Pattern matched multiple keys")
    135 key = keys[0]
    137 if not silent:

OptionError: Pattern matched multiple keys

警告

使用这种简写形式可能会导致您的代码在未来版本中添加了具有类似名称的新选项时出现故障。

可用选项#

您可以使用 describe_option() 获取可用选项及其描述的列表。当不带参数调用时,describe_option() 将打印所有可用选项的描述。

In [11]: pd.describe_option()
compute.use_bottleneck : bool
    Use the bottleneck library to accelerate if it is installed,
    the default is True
    Valid values: False,True
    [default: True] [currently: True]
compute.use_numba : bool
    Use the numba engine option for select operations if it is installed,
    the default is False
    Valid values: False,True
    [default: False] [currently: False]
compute.use_numexpr : bool
    Use the numexpr library to accelerate computation if it is installed,
    the default is True
    Valid values: False,True
    [default: True] [currently: True]
display.chop_threshold : float or None
    if set to a float value, all float values smaller than the given threshold
    will be displayed as exactly 0 by repr and friends.
    [default: None] [currently: None]
display.colheader_justify : 'left'/'right'
    Controls the justification of column headers. used by DataFrameFormatter.
    [default: right] [currently: right]
display.date_dayfirst : boolean
    When True, prints and parses dates with the day first, eg 20/01/2005
    [default: False] [currently: False]
display.date_yearfirst : boolean
    When True, prints and parses dates with the year first, eg 2005/01/20
    [default: False] [currently: False]
display.encoding : str/unicode
    Defaults to the detected encoding of the console.
    Specifies the encoding to be used for strings returned by to_string,
    these are generally strings meant to be displayed on the console.
    [default: utf-8] [currently: utf8]
display.expand_frame_repr : boolean
    Whether to print out the full DataFrame repr for wide DataFrames across
    multiple lines, `max_columns` is still respected, but the output will
    wrap-around across multiple "pages" if its width exceeds `display.width`.
    [default: True] [currently: True]
display.float_format : callable
    The callable should accept a floating point number and return
    a string with the desired format of the number. This is used
    in some places like SeriesFormatter.
    See formats.format.EngFormatter for an example.
    [default: None] [currently: None]
display.html.border : int
    A ``border=value`` attribute is inserted in the ``<table>`` tag
    for the DataFrame HTML repr.
    [default: 1] [currently: 1]
display.html.table_schema : boolean
    Whether to publish a Table Schema representation for frontends
    that support it.
    (default: False)
    [default: False] [currently: False]
display.html.use_mathjax : boolean
    When True, Jupyter notebook will process table contents using MathJax,
    rendering mathematical expressions enclosed by the dollar symbol.
    (default: True)
    [default: True] [currently: True]
display.large_repr : 'truncate'/'info'
    For DataFrames exceeding max_rows/max_cols, the repr (and HTML repr) can
    show a truncated table, or switch to the view from
    df.info() (the behaviour in earlier versions of pandas).
    [default: truncate] [currently: truncate]
display.max_categories : int
    This sets the maximum number of categories pandas should output when
    printing out a `Categorical` or a Series of dtype "category".
    [default: 8] [currently: 8]
display.max_columns : int
    If max_cols is exceeded, switch to truncate view. Depending on
    `large_repr`, objects are either centrally truncated or printed as
    a summary view. 'None' value means unlimited.

    In case python/IPython is running in a terminal and `large_repr`
    equals 'truncate' this can be set to 0 or None and pandas will auto-detect
    the width of the terminal and print a truncated object which fits
    the screen width. The IPython notebook, IPython qtconsole, or IDLE
    do not run in a terminal and hence it is not possible to do
    correct auto-detection and defaults to 20.
    [default: 0] [currently: 0]
display.max_colwidth : int or None
    The maximum width in characters of a column in the repr of
    a pandas data structure. When the column overflows, a "..."
    placeholder is embedded in the output. A 'None' value means unlimited.
    [default: 50] [currently: 50]
display.max_dir_items : int
    The number of items that will be added to `dir(...)`. 'None' value means
    unlimited. Because dir is cached, changing this option will not immediately
    affect already existing dataframes until a column is deleted or added.

    This is for instance used to suggest columns from a dataframe to tab
    completion.
    [default: 100] [currently: 100]
display.max_info_columns : int
    max_info_columns is used in DataFrame.info method to decide if
    per column information will be printed.
    [default: 100] [currently: 100]
display.max_info_rows : int
    df.info() will usually show null-counts for each column.
    For large frames this can be quite slow. max_info_rows and max_info_cols
    limit this null check only to frames with smaller dimensions than
    specified.
    [default: 1690785] [currently: 1690785]
display.max_rows : int
    If max_rows is exceeded, switch to truncate view. Depending on
    `large_repr`, objects are either centrally truncated or printed as
    a summary view. 'None' value means unlimited.

    In case python/IPython is running in a terminal and `large_repr`
    equals 'truncate' this can be set to 0 and pandas will auto-detect
    the height of the terminal and print a truncated object which fits
    the screen height. The IPython notebook, IPython qtconsole, or
    IDLE do not run in a terminal and hence it is not possible to do
    correct auto-detection.
    [default: 60] [currently: 60]
display.max_seq_items : int or None
    When pretty-printing a long sequence, no more then `max_seq_items`
    will be printed. If items are omitted, they will be denoted by the
    addition of "..." to the resulting string.

    If set to None, the number of items to be printed is unlimited.
    [default: 100] [currently: 100]
display.memory_usage : bool, string or None
    This specifies if the memory usage of a DataFrame should be displayed when
    df.info() is called. Valid values True,False,'deep'
    [default: True] [currently: True]
display.min_rows : int
    The numbers of rows to show in a truncated view (when `max_rows` is
    exceeded). Ignored when `max_rows` is set to None or 0. When set to
    None, follows the value of `max_rows`.
    [default: 10] [currently: 10]
display.multi_sparse : boolean
    "sparsify" MultiIndex display (don't display repeated
    elements in outer levels within groups)
    [default: True] [currently: True]
display.notebook_repr_html : boolean
    When True, IPython notebook will use html representation for
    pandas objects (if it is available).
    [default: True] [currently: True]
display.pprint_nest_depth : int
    Controls the number of nested levels to process when pretty-printing
    [default: 3] [currently: 3]
display.precision : int
    Floating point output precision in terms of number of places after the
    decimal, for regular formatting as well as scientific notation. Similar
    to ``precision`` in :meth:`numpy.set_printoptions`.
    [default: 6] [currently: 6]
display.show_dimensions : boolean or 'truncate'
    Whether to print out dimensions at the end of DataFrame repr.
    If 'truncate' is specified, only print out the dimensions if the
    frame is truncated (e.g. not display all rows and/or columns)
    [default: truncate] [currently: truncate]
display.unicode.ambiguous_as_wide : boolean
    Whether to use the Unicode East Asian Width to calculate the display text
    width.
    Enabling this may affect to the performance (default: False)
    [default: False] [currently: False]
display.unicode.east_asian_width : boolean
    Whether to use the Unicode East Asian Width to calculate the display text
    width.
    Enabling this may affect to the performance (default: False)
    [default: False] [currently: False]
display.width : int
    Width of the display in characters. In case python/IPython is running in
    a terminal this can be set to None and pandas will correctly auto-detect
    the width.
    Note that the IPython notebook, IPython qtconsole, or IDLE do not run in a
    terminal and hence it is not possible to correctly detect the width.
    [default: 80] [currently: 80]
future.infer_string Whether to infer sequence of str objects as pyarrow string dtype, which will be the default in pandas 3.0 (at which point this option will be deprecated).
    [default: False] [currently: False]
future.no_silent_downcasting Whether to opt-in to the future behavior which will *not* silently downcast results from Series and DataFrame `where`, `mask`, and `clip` methods. Silent downcasting will be removed in pandas 3.0 (at which point this option will be deprecated).
    [default: False] [currently: False]
io.excel.ods.reader : string
    The default Excel reader engine for 'ods' files. Available options:
    auto, odf, calamine.
    [default: auto] [currently: auto]
io.excel.ods.writer : string
    The default Excel writer engine for 'ods' files. Available options:
    auto, odf.
    [default: auto] [currently: auto]
io.excel.xls.reader : string
    The default Excel reader engine for 'xls' files. Available options:
    auto, xlrd, calamine.
    [default: auto] [currently: auto]
io.excel.xlsb.reader : string
    The default Excel reader engine for 'xlsb' files. Available options:
    auto, pyxlsb, calamine.
    [default: auto] [currently: auto]
io.excel.xlsm.reader : string
    The default Excel reader engine for 'xlsm' files. Available options:
    auto, xlrd, openpyxl, calamine.
    [default: auto] [currently: auto]
io.excel.xlsm.writer : string
    The default Excel writer engine for 'xlsm' files. Available options:
    auto, openpyxl.
    [default: auto] [currently: auto]
io.excel.xlsx.reader : string
    The default Excel reader engine for 'xlsx' files. Available options:
    auto, xlrd, openpyxl, calamine.
    [default: auto] [currently: auto]
io.excel.xlsx.writer : string
    The default Excel writer engine for 'xlsx' files. Available options:
    auto, openpyxl, xlsxwriter.
    [default: auto] [currently: auto]
io.hdf.default_format : format
    default format writing format, if None, then
    put will default to 'fixed' and append will default to 'table'
    [default: None] [currently: None]
io.hdf.dropna_table : boolean
    drop ALL nan rows when appending to a table
    [default: False] [currently: False]
io.parquet.engine : string
    The default parquet reader/writer engine. Available options:
    'auto', 'pyarrow', 'fastparquet', the default is 'auto'
    [default: auto] [currently: auto]
io.sql.engine : string
    The default sql reader/writer engine. Available options:
    'auto', 'sqlalchemy', the default is 'auto'
    [default: auto] [currently: auto]
mode.chained_assignment : string
    Raise an exception, warn, or no action if trying to use chained assignment,
    The default is warn
    [default: warn] [currently: warn]
mode.copy_on_write : bool
    Use new copy-view behaviour using Copy-on-Write. Defaults to False,
    unless overridden by the 'PANDAS_COPY_ON_WRITE' environment variable
    (if set to "1" for True, needs to be set before pandas is imported).
    [default: False] [currently: False]
mode.data_manager : string
    Internal data manager type; can be "block" or "array". Defaults to "block",
    unless overridden by the 'PANDAS_DATA_MANAGER' environment variable (needs
    to be set before pandas is imported).
    [default: block] [currently: block]
    (Deprecated, use `` instead.)
mode.sim_interactive : boolean
    Whether to simulate interactive mode for purposes of testing
    [default: False] [currently: False]
mode.string_storage : string
    The default storage for StringDtype. This option is ignored if
    ``future.infer_string`` is set to True.
    [default: python] [currently: python]
mode.use_inf_as_na : boolean
    True means treat None, NaN, INF, -INF as NA (old way),
    False means None and NaN are null, but INF, -INF are not NA
    (new way).

    This option is deprecated in pandas 2.1.0 and will be removed in 3.0.
    [default: False] [currently: False]
    (Deprecated, use `` instead.)
plotting.backend : str
    The plotting backend to use. The default value is "matplotlib", the
    backend provided with pandas. Other backends can be specified by
    providing the name of the module that implements the backend.
    [default: matplotlib] [currently: matplotlib]
plotting.matplotlib.register_converters : bool or 'auto'.
    Whether to register converters with matplotlib's units registry for
    dates, times, datetimes, and Periods. Toggling to False will remove
    the converters, restoring any converters that pandas overwrote.
    [default: auto] [currently: auto]
styler.format.decimal : str
    The character representation for the decimal separator for floats and complex.
    [default: .] [currently: .]
styler.format.escape : str, optional
    Whether to escape certain characters according to the given context; html or latex.
    [default: None] [currently: None]
styler.format.formatter : str, callable, dict, optional
    A formatter object to be used as default within ``Styler.format``.
    [default: None] [currently: None]
styler.format.na_rep : str, optional
    The string representation for values identified as missing.
    [default: None] [currently: None]
styler.format.precision : int
    The precision for floats and complex numbers.
    [default: 6] [currently: 6]
styler.format.thousands : str, optional
    The character representation for thousands separator for floats, int and complex.
    [default: None] [currently: None]
styler.html.mathjax : bool
    If False will render special CSS classes to table attributes that indicate Mathjax
    will not be used in Jupyter Notebook.
    [default: True] [currently: True]
styler.latex.environment : str
    The environment to replace ``\begin{table}``. If "longtable" is used results
    in a specific longtable environment format.
    [default: None] [currently: None]
styler.latex.hrules : bool
    Whether to add horizontal rules on top and bottom and below the headers.
    [default: False] [currently: False]
styler.latex.multicol_align : {"r", "c", "l", "naive-l", "naive-r"}
    The specifier for horizontal alignment of sparsified LaTeX multicolumns. Pipe
    decorators can also be added to non-naive values to draw vertical
    rules, e.g. "\|r" will draw a rule on the left side of right aligned merged cells.
    [default: r] [currently: r]
styler.latex.multirow_align : {"c", "t", "b"}
    The specifier for vertical alignment of sparsified LaTeX multirows.
    [default: c] [currently: c]
styler.render.encoding : str
    The encoding used for output HTML and LaTeX files.
    [default: utf-8] [currently: utf-8]
styler.render.max_columns : int, optional
    The maximum number of columns that will be rendered. May still be reduced to
    satisfy ``max_elements``, which takes precedence.
    [default: None] [currently: None]
styler.render.max_elements : int
    The maximum number of data-cell (<td>) elements that will be rendered before
    trimming will occur over columns, rows or both if needed.
    [default: 262144] [currently: 262144]
styler.render.max_rows : int, optional
    The maximum number of rows that will be rendered. May still be reduced to
    satisfy ``max_elements``, which takes precedence.
    [default: None] [currently: None]
styler.render.repr : str
    Determine which output to use in Jupyter Notebook in {"html", "latex"}.
    [default: html] [currently: html]
styler.sparse.columns : bool
    Whether to sparsify the display of hierarchical columns. Setting to False will
    display each explicit level element in a hierarchical key for each column.
    [default: True] [currently: True]
styler.sparse.index : bool
    Whether to sparsify the display of a hierarchical index. Setting to False will
    display each explicit level element in a hierarchical key for each row.
    [default: True] [currently: True]

获取和设置选项#

如上所述,get_option()set_option() 可从 pandas 命名空间获得。要更改选项,请调用 set_option('option regex', new_value)

In [12]: pd.get_option("mode.sim_interactive")
Out[12]: False

In [13]: pd.set_option("mode.sim_interactive", True)

In [14]: pd.get_option("mode.sim_interactive")
Out[14]: True

注意

选项 'mode.sim_interactive' 主要用于调试目的。

您可以使用 reset_option() 将设置恢复为默认值。

In [15]: pd.get_option("display.max_rows")
Out[15]: 60

In [16]: pd.set_option("display.max_rows", 999)

In [17]: pd.get_option("display.max_rows")
Out[17]: 999

In [18]: pd.reset_option("display.max_rows")

In [19]: pd.get_option("display.max_rows")
Out[19]: 60

也可以使用正则表达式一次重置多个选项。

In [20]: pd.reset_option("^display")

option_context() 上下文管理器已通过顶级 API 公开,允许您使用给定的选项值执行代码。当您退出 with 块时,选项值会自动恢复。

In [21]: with pd.option_context("display.max_rows", 10, "display.max_columns", 5):
   ....:     print(pd.get_option("display.max_rows"))
   ....:     print(pd.get_option("display.max_columns"))
   ....: 
10
5

In [22]: print(pd.get_option("display.max_rows"))
60

In [23]: print(pd.get_option("display.max_columns"))
0

在 Python/IPython 环境中设置启动选项#

使用 Python/IPython 环境的启动脚本导入 pandas 并设置选项,可以提高使用 pandas 的效率。为此,请在所需配置文件的启动目录中创建一个 .py.ipy 脚本。可以在默认 IPython 配置文件中的启动文件夹中找到一个示例,如下所示:

$IPYTHONDIR/profile_default/startup

有关更多信息,请参阅 IPython 文档。下面显示了 pandas 的示例启动脚本。

import pandas as pd

pd.set_option("display.max_rows", 999)
pd.set_option("display.precision", 5)

常用选项#

以下是更常用的显示选项的演示。

display.max_rowsdisplay.max_columns 设置在漂亮打印框架时显示的最大行数和列数。截断的行将用省略号替换。

In [24]: df = pd.DataFrame(np.random.randn(7, 2))

In [25]: pd.set_option("display.max_rows", 7)

In [26]: df
Out[26]: 
          0         1
0  0.469112 -0.282863
1 -1.509059 -1.135632
2  1.212112 -0.173215
3  0.119209 -1.044236
4 -0.861849 -2.104569
5 -0.494929  1.071804
6  0.721555 -0.706771

In [27]: pd.set_option("display.max_rows", 5)

In [28]: df
Out[28]: 
           0         1
0   0.469112 -0.282863
1  -1.509059 -1.135632
..       ...       ...
5  -0.494929  1.071804
6   0.721555 -0.706771

[7 rows x 2 columns]

In [29]: pd.reset_option("display.max_rows")

一旦超过 display.max_rowsdisplay.min_rows 选项将确定在截断的 repr 中显示多少行。

In [30]: pd.set_option("display.max_rows", 8)

In [31]: pd.set_option("display.min_rows", 4)

# below max_rows -> all rows shown
In [32]: df = pd.DataFrame(np.random.randn(7, 2))

In [33]: df
Out[33]: 
          0         1
0 -1.039575  0.271860
1 -0.424972  0.567020
2  0.276232 -1.087401
3 -0.673690  0.113648
4 -1.478427  0.524988
5  0.404705  0.577046
6 -1.715002 -1.039268

# above max_rows -> only min_rows (4) rows shown
In [34]: df = pd.DataFrame(np.random.randn(9, 2))

In [35]: df
Out[35]: 
           0         1
0  -0.370647 -1.157892
1  -1.344312  0.844885
..       ...       ...
7   0.276662 -0.472035
8  -0.013960 -0.362543

[9 rows x 2 columns]

In [36]: pd.reset_option("display.max_rows")

In [37]: pd.reset_option("display.min_rows")

display.expand_frame_repr 允许 DataFrame 的表示跨越页面,在所有列上进行包装。

In [38]: df = pd.DataFrame(np.random.randn(5, 10))

In [39]: pd.set_option("expand_frame_repr", True)

In [40]: df
Out[40]: 
          0         1         2  ...         7         8         9
0 -0.006154 -0.923061  0.895717  ...  1.340309 -1.170299 -0.226169
1  0.410835  0.813850  0.132003  ... -1.436737 -1.413681  1.607920
2  1.024180  0.569605  0.875906  ... -0.078638  0.545952 -1.219217
3 -1.226825  0.769804 -1.281247  ...  0.341734  0.959726 -1.110336
4 -0.619976  0.149748 -0.732339  ...  0.301624 -2.179861 -1.369849

[5 rows x 10 columns]

In [41]: pd.set_option("expand_frame_repr", False)

In [42]: df
Out[42]: 
          0         1         2         3         4         5         6         7         8         9
0 -0.006154 -0.923061  0.895717  0.805244 -1.206412  2.565646  1.431256  1.340309 -1.170299 -0.226169
1  0.410835  0.813850  0.132003 -0.827317 -0.076467 -1.187678  1.130127 -1.436737 -1.413681  1.607920
2  1.024180  0.569605  0.875906 -2.211372  0.974466 -2.006747 -0.410001 -0.078638  0.545952 -1.219217
3 -1.226825  0.769804 -1.281247 -0.727707 -0.121306 -0.097883  0.695775  0.341734  0.959726 -1.110336
4 -0.619976  0.149748 -0.732339  0.687738  0.176444  0.403310 -0.154951  0.301624 -2.179861 -1.369849

In [43]: pd.reset_option("expand_frame_repr")

display.large_repr 显示超过 max_columnsmax_rowsDataFrame 作为截断的框架或摘要。

In [44]: df = pd.DataFrame(np.random.randn(10, 10))

In [45]: pd.set_option("display.max_rows", 5)

In [46]: pd.set_option("large_repr", "truncate")

In [47]: df
Out[47]: 
           0         1         2  ...         7         8         9
0  -0.954208  1.462696 -1.743161  ...  0.995761  2.396780  0.014871
1   3.357427 -0.317441 -1.236269  ...  0.380396  0.084844  0.432390
..       ...       ...       ...  ...       ...       ...       ...
8  -0.303421 -0.858447  0.306996  ...  0.476720  0.473424 -0.242861
9  -0.014805 -0.284319  0.650776  ...  1.613616  0.464000  0.227371

[10 rows x 10 columns]

In [48]: pd.set_option("large_repr", "info")

In [49]: df
Out[49]: 
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 10 entries, 0 to 9
Data columns (total 10 columns):
 #   Column  Non-Null Count  Dtype  
---  ------  --------------  -----  
 0   0       10 non-null     float64
 1   1       10 non-null     float64
 2   2       10 non-null     float64
 3   3       10 non-null     float64
 4   4       10 non-null     float64
 5   5       10 non-null     float64
 6   6       10 non-null     float64
 7   7       10 non-null     float64
 8   8       10 non-null     float64
 9   9       10 non-null     float64
dtypes: float64(10)
memory usage: 928.0 bytes

In [50]: pd.reset_option("large_repr")

In [51]: pd.reset_option("display.max_rows")

display.max_colwidth 设置列的最大宽度。长度等于或超过此长度的单元格将被截断,并用省略号表示。

In [52]: df = pd.DataFrame(
   ....:     np.array(
   ....:         [
   ....:             ["foo", "bar", "bim", "uncomfortably long string"],
   ....:             ["horse", "cow", "banana", "apple"],
   ....:         ]
   ....:     )
   ....: )
   ....: 

In [53]: pd.set_option("max_colwidth", 40)

In [54]: df
Out[54]: 
       0    1       2                          3
0    foo  bar     bim  uncomfortably long string
1  horse  cow  banana                      apple

In [55]: pd.set_option("max_colwidth", 6)

In [56]: df
Out[56]: 
       0    1      2      3
0    foo  bar    bim  un...
1  horse  cow  ba...  apple

In [57]: pd.reset_option("max_colwidth")

display.max_info_columns 设置调用 info() 时显示的列数阈值。

In [58]: df = pd.DataFrame(np.random.randn(10, 10))

In [59]: pd.set_option("max_info_columns", 11)

In [60]: df.info()
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 10 entries, 0 to 9
Data columns (total 10 columns):
 #   Column  Non-Null Count  Dtype  
---  ------  --------------  -----  
 0   0       10 non-null     float64
 1   1       10 non-null     float64
 2   2       10 non-null     float64
 3   3       10 non-null     float64
 4   4       10 non-null     float64
 5   5       10 non-null     float64
 6   6       10 non-null     float64
 7   7       10 non-null     float64
 8   8       10 non-null     float64
 9   9       10 non-null     float64
dtypes: float64(10)
memory usage: 928.0 bytes

In [61]: pd.set_option("max_info_columns", 5)

In [62]: df.info()
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 10 entries, 0 to 9
Columns: 10 entries, 0 to 9
dtypes: float64(10)
memory usage: 928.0 bytes

In [63]: pd.reset_option("max_info_columns")

display.max_info_rows: info() 通常会显示每个列的空值计数。对于大型 DataFrame 来说,这可能非常慢。 max_info_rowsmax_info_cols 将此空值检查限制在指定的行和列。 info() 的关键字参数 show_counts=True 将覆盖此设置。

In [64]: df = pd.DataFrame(np.random.choice([0, 1, np.nan], size=(10, 10)))

In [65]: df
Out[65]: 
     0    1    2    3    4    5    6    7    8    9
0  0.0  NaN  1.0  NaN  NaN  0.0  NaN  0.0  NaN  1.0
1  1.0  NaN  1.0  1.0  1.0  1.0  NaN  0.0  0.0  NaN
2  0.0  NaN  1.0  0.0  0.0  NaN  NaN  NaN  NaN  0.0
3  NaN  NaN  NaN  0.0  1.0  1.0  NaN  1.0  NaN  1.0
4  0.0  NaN  NaN  NaN  0.0  NaN  NaN  NaN  1.0  0.0
5  0.0  1.0  1.0  1.0  1.0  0.0  NaN  NaN  1.0  0.0
6  1.0  1.0  1.0  NaN  1.0  NaN  1.0  0.0  NaN  NaN
7  0.0  0.0  1.0  0.0  1.0  0.0  1.0  1.0  0.0  NaN
8  NaN  NaN  NaN  0.0  NaN  NaN  NaN  NaN  1.0  NaN
9  0.0  NaN  0.0  NaN  NaN  0.0  NaN  1.0  1.0  0.0

In [66]: pd.set_option("max_info_rows", 11)

In [67]: df.info()
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 10 entries, 0 to 9
Data columns (total 10 columns):
 #   Column  Non-Null Count  Dtype  
---  ------  --------------  -----  
 0   0       8 non-null      float64
 1   1       3 non-null      float64
 2   2       7 non-null      float64
 3   3       6 non-null      float64
 4   4       7 non-null      float64
 5   5       6 non-null      float64
 6   6       2 non-null      float64
 7   7       6 non-null      float64
 8   8       6 non-null      float64
 9   9       6 non-null      float64
dtypes: float64(10)
memory usage: 928.0 bytes

In [68]: pd.set_option("max_info_rows", 5)

In [69]: df.info()
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 10 entries, 0 to 9
Data columns (total 10 columns):
 #   Column  Dtype  
---  ------  -----  
 0   0       float64
 1   1       float64
 2   2       float64
 3   3       float64
 4   4       float64
 5   5       float64
 6   6       float64
 7   7       float64
 8   8       float64
 9   9       float64
dtypes: float64(10)
memory usage: 928.0 bytes

In [70]: pd.reset_option("max_info_rows")

display.precision 设置输出显示精度(以小数位数表示)。

In [71]: df = pd.DataFrame(np.random.randn(5, 5))

In [72]: pd.set_option("display.precision", 7)

In [73]: df
Out[73]: 
           0          1          2          3          4
0 -1.1506406 -0.7983341 -0.5576966  0.3813531  1.3371217
1 -1.5310949  1.3314582 -0.5713290 -0.0266708 -1.0856630
2 -1.1147378 -0.0582158 -0.4867681  1.6851483  0.1125723
3 -1.4953086  0.8984347 -0.1482168 -1.5960698  0.1596530
4  0.2621358  0.0362196  0.1847350 -0.2550694 -0.2710197

In [74]: pd.set_option("display.precision", 4)

In [75]: df
Out[75]: 
        0       1       2       3       4
0 -1.1506 -0.7983 -0.5577  0.3814  1.3371
1 -1.5311  1.3315 -0.5713 -0.0267 -1.0857
2 -1.1147 -0.0582 -0.4868  1.6851  0.1126
3 -1.4953  0.8984 -0.1482 -1.5961  0.1597
4  0.2621  0.0362  0.1847 -0.2551 -0.2710

display.chop_threshold 设置在显示 SeriesDataFrame 时将舍入到零的阈值。此设置不会更改数字存储的精度。

In [76]: df = pd.DataFrame(np.random.randn(6, 6))

In [77]: pd.set_option("chop_threshold", 0)

In [78]: df
Out[78]: 
        0       1       2       3       4       5
0  1.2884  0.2946 -1.1658  0.8470 -0.6856  0.6091
1 -0.3040  0.6256 -0.0593  0.2497  1.1039 -1.0875
2  1.9980 -0.2445  0.1362  0.8863 -1.3507 -0.8863
3 -1.0133  1.9209 -0.3882 -2.3144  0.6655  0.4026
4  0.3996 -1.7660  0.8504  0.3881  0.9923  0.7441
5 -0.7398 -1.0549 -0.1796  0.6396  1.5850  1.9067

In [79]: pd.set_option("chop_threshold", 0.5)

In [80]: df
Out[80]: 
        0       1       2       3       4       5
0  1.2884  0.0000 -1.1658  0.8470 -0.6856  0.6091
1  0.0000  0.6256  0.0000  0.0000  1.1039 -1.0875
2  1.9980  0.0000  0.0000  0.8863 -1.3507 -0.8863
3 -1.0133  1.9209  0.0000 -2.3144  0.6655  0.0000
4  0.0000 -1.7660  0.8504  0.0000  0.9923  0.7441
5 -0.7398 -1.0549  0.0000  0.6396  1.5850  1.9067

In [81]: pd.reset_option("chop_threshold")

display.colheader_justify 控制标题的对齐方式。选项包括 'right''left'

In [82]: df = pd.DataFrame(
   ....:     np.array([np.random.randn(6), np.random.randint(1, 9, 6) * 0.1, np.zeros(6)]).T,
   ....:     columns=["A", "B", "C"],
   ....:     dtype="float",
   ....: )
   ....: 

In [83]: pd.set_option("colheader_justify", "right")

In [84]: df
Out[84]: 
        A    B    C
0  0.1040  0.1  0.0
1  0.1741  0.5  0.0
2 -0.4395  0.4  0.0
3 -0.7413  0.8  0.0
4 -0.0797  0.4  0.0
5 -0.9229  0.3  0.0

In [85]: pd.set_option("colheader_justify", "left")

In [86]: df
Out[86]: 
   A       B    C  
0  0.1040  0.1  0.0
1  0.1741  0.5  0.0
2 -0.4395  0.4  0.0
3 -0.7413  0.8  0.0
4 -0.0797  0.4  0.0
5 -0.9229  0.3  0.0

In [87]: pd.reset_option("colheader_justify")

数字格式化#

pandas 还允许您设置数字在控制台中的显示方式。此选项不是通过 set_options API 设置的。

使用 set_eng_float_format 函数更改 pandas 对象的浮点数格式以生成特定格式。

In [88]: import numpy as np

In [89]: pd.set_eng_float_format(accuracy=3, use_eng_prefix=True)

In [90]: s = pd.Series(np.random.randn(5), index=["a", "b", "c", "d", "e"])

In [91]: s / 1.0e3
Out[91]: 
a    303.638u
b   -721.084u
c   -622.696u
d    648.250u
e     -1.945m
dtype: float64

In [92]: s / 1.0e6
Out[92]: 
a    303.638n
b   -721.084n
c   -622.696n
d    648.250n
e     -1.945u
dtype: float64

使用 round() 来专门控制单个 DataFrame 的舍入。

Unicode 格式化#

警告

启用此选项会影响 DataFrame 和 Series 打印的性能(大约慢两倍)。仅在实际需要时使用。

一些东亚国家使用 Unicode 字符,其宽度对应于两个拉丁字符。如果 DataFrame 或 Series 包含这些字符,默认输出模式可能无法正确对齐它们。

In [93]: df = pd.DataFrame({"国籍": ["UK", "日本"], "名前": ["Alice", "しのぶ"]})

In [94]: df
Out[94]: 
   国籍     名前
0  UK  Alice
1  日本    しのぶ

启用 display.unicode.east_asian_width 允许 pandas 检查每个字符的“东亚宽度”属性。通过将此选项设置为 True,可以正确对齐这些字符。但是,这会导致比标准 len 函数更长的渲染时间。

In [95]: pd.set_option("display.unicode.east_asian_width", True)

In [96]: df
Out[96]: 
   国籍    名前
0    UK   Alice
1  日本  しのぶ

此外,宽度为“模糊”的 Unicode 字符可以是 1 或 2 个字符宽,具体取决于终端设置或编码。选项 display.unicode.ambiguous_as_wide 可用于处理这种歧义。

默认情况下,将“模糊”字符的宽度(例如以下示例中的“¡”(倒置感叹号))视为 1。

In [97]: df = pd.DataFrame({"a": ["xxx", "¡¡"], "b": ["yyy", "¡¡"]})

In [98]: df
Out[98]: 
     a    b
0  xxx  yyy
1   ¡¡   ¡¡

启用 display.unicode.ambiguous_as_wide 会使 pandas 将这些字符的宽度解释为 2。(请注意,此选项仅在启用 display.unicode.east_asian_width 时有效。)

但是,如果在您的终端中错误地设置了此选项,会导致这些字符对齐错误。

In [99]: pd.set_option("display.unicode.ambiguous_as_wide", True)

In [100]: df
Out[100]: 
      a     b
0   xxx   yyy
1  ¡¡  ¡¡

表格模式显示#

DataFrameSeries 默认情况下会发布表格模式表示。这可以通过 display.html.table_schema 选项在全局范围内启用。

In [101]: pd.set_option("display.html.table_schema", True)

仅序列化和发布 'display.max_rows'