标签归档：csv-import

知识问答

熊猫read_csv并使用usecols过滤列

问题：熊猫read_csv并使用usecols过滤列

我有一个csv文件，pandas.read_csv当我使用过滤列usecols并使用多个索引时，该文件输入不正确。

import pandas as pd
csv = r"""dummy,date,loc,x
   bar,20090101,a,1
   bar,20090102,a,3
   bar,20090103,a,5
   bar,20090101,b,1
   bar,20090102,b,3
   bar,20090103,b,5"""

f = open('foo.csv', 'w')
f.write(csv)
f.close()

df1 = pd.read_csv('foo.csv',
        header=0,
        names=["dummy", "date", "loc", "x"], 
        index_col=["date", "loc"], 
        usecols=["dummy", "date", "loc", "x"],
        parse_dates=["date"])
print df1

# Ignore the dummy columns
df2 = pd.read_csv('foo.csv', 
        index_col=["date", "loc"], 
        usecols=["date", "loc", "x"], # <----------- Changed
        parse_dates=["date"],
        header=0,
        names=["dummy", "date", "loc", "x"])
print df2

我希望df1和df2除了丢失的虚拟列外应该相同，但是这些列的标签错误。日期也被解析为日期。 

In [118]: %run test.py
               dummy  x
date       loc
2009-01-01 a     bar  1
2009-01-02 a     bar  3
2009-01-03 a     bar  5
2009-01-01 b     bar  1
2009-01-02 b     bar  3
2009-01-03 b     bar  5
              date
date loc
a    1    20090101
     3    20090102
     5    20090103
b    1    20090101
     3    20090102
     5    20090103

使用列号而不是名称给我同样的问题。我可以通过在read_csv步骤之后删除虚拟列来解决此问题，但是我试图了解出了什么问题。我正在使用熊猫0.10.1。

编辑：修复错误的标头用法。

点击查看英文原文

I have a csv file which isn’t coming in correctly with pandas.read_csv when I filter the columns with usecols and use multiple indexes.

import pandas as pd
csv = r"""dummy,date,loc,x
   bar,20090101,a,1
   bar,20090102,a,3
   bar,20090103,a,5
   bar,20090101,b,1
   bar,20090102,b,3
   bar,20090103,b,5"""

f = open('foo.csv', 'w')
f.write(csv)
f.close()

df1 = pd.read_csv('foo.csv',
        header=0,
        names=["dummy", "date", "loc", "x"], 
        index_col=["date", "loc"], 
        usecols=["dummy", "date", "loc", "x"],
        parse_dates=["date"])
print df1

# Ignore the dummy columns
df2 = pd.read_csv('foo.csv', 
        index_col=["date", "loc"], 
        usecols=["date", "loc", "x"], # <----------- Changed
        parse_dates=["date"],
        header=0,
        names=["dummy", "date", "loc", "x"])
print df2

I expect that df1 and df2 should be the same except for the missing dummy column, but the columns come in mislabeled. Also the date is getting parsed as a date. 

In [118]: %run test.py
               dummy  x
date       loc
2009-01-01 a     bar  1
2009-01-02 a     bar  3
2009-01-03 a     bar  5
2009-01-01 b     bar  1
2009-01-02 b     bar  3
2009-01-03 b     bar  5
              date
date loc
a    1    20090101
     3    20090102
     5    20090103
b    1    20090101
     3    20090102
     5    20090103

Using column numbers instead of names give me the same problem. I can workaround the issue by dropping the dummy column after the read_csv step, but I’m trying to understand what is going wrong. I’m using pandas 0.10.1.

edit: fixed bad header usage.

回答 0

@chip的答案完全错过了两个关键字参数的含义。

  • 名称是只在必要时有没有头和要指定使用的列名，而不是整数索引等参数。
  • usecols应该在将整个DataFrame读入内存之前提供过滤器；如果使用得当，则读取后永远不需要删除列。

此解决方案纠正了这些怪异现象：

import pandas as pd
from StringIO import StringIO

csv = r"""dummy,date,loc,x
bar,20090101,a,1
bar,20090102,a,3
bar,20090103,a,5
bar,20090101,b,1
bar,20090102,b,3
bar,20090103,b,5"""

df = pd.read_csv(StringIO(csv),
        header=0,
        index_col=["date", "loc"], 
        usecols=["date", "loc", "x"],
        parse_dates=["date"])

这给了我们：

                x
date       loc
2009-01-01 a    1
2009-01-02 a    3
2009-01-03 a    5
2009-01-01 b    1
2009-01-02 b    3
2009-01-03 b    5
点击查看英文原文

The answer by @chip completely misses the point of two keyword arguments.

  • names is only necessary when there is no header and you want to specify other arguments using column names rather than integer indices.
  • usecols is supposed to provide a filter before reading the whole DataFrame into memory; if used properly, there should never be a need to delete columns after reading.

This solution corrects those oddities:

import pandas as pd
from StringIO import StringIO

csv = r"""dummy,date,loc,x
bar,20090101,a,1
bar,20090102,a,3
bar,20090103,a,5
bar,20090101,b,1
bar,20090102,b,3
bar,20090103,b,5"""

df = pd.read_csv(StringIO(csv),
        header=0,
        index_col=["date", "loc"], 
        usecols=["date", "loc", "x"],
        parse_dates=["date"])

Which gives us:

                x
date       loc
2009-01-01 a    1
2009-01-02 a    3
2009-01-03 a    5
2009-01-01 b    1
2009-01-02 b    3
2009-01-03 b    5

回答 1

这段代码实现了您想要的—也很奇怪，甚至有很多错误：

我观察到它在以下情况下有效：

a）您指定index_col相关。到您实际使用的列数-因此在此示例中为三列，而不是四列（dummy从此开始滴下并开始计数）

b）相同 parse_dates

c）并非usecols出于明显原因；）

d）在这里我改编了names以反映此行为

import pandas as pd
from StringIO import StringIO

csv = """dummy,date,loc,x
bar,20090101,a,1
bar,20090102,a,3
bar,20090103,a,5
bar,20090101,b,1
bar,20090102,b,3
bar,20090103,b,5
"""

df = pd.read_csv(StringIO(csv),
        index_col=[0,1],
        usecols=[1,2,3], 
        parse_dates=[0],
        header=0,
        names=["date", "loc", "", "x"])

print df

哪个打印

                x
date       loc   
2009-01-01 a    1
2009-01-02 a    3
2009-01-03 a    5
2009-01-01 b    1
2009-01-02 b    3
2009-01-03 b    5
点击查看英文原文

This code achieves what you want — also its weird and certainly buggy:

I observed that it works when:

a) you specify the index_col rel. to the number of columns you really use — so its three columns in this example, not four (you drop dummy and start counting from then onwards)

b) same for parse_dates

c) not so for usecols ;) for obvious reasons

d) here I adapted the names to mirror this behaviour

import pandas as pd
from StringIO import StringIO

csv = """dummy,date,loc,x
bar,20090101,a,1
bar,20090102,a,3
bar,20090103,a,5
bar,20090101,b,1
bar,20090102,b,3
bar,20090103,b,5
"""

df = pd.read_csv(StringIO(csv),
        index_col=[0,1],
        usecols=[1,2,3], 
        parse_dates=[0],
        header=0,
        names=["date", "loc", "", "x"])

print df

which prints

                x
date       loc   
2009-01-01 a    1
2009-01-02 a    3
2009-01-03 a    5
2009-01-01 b    1
2009-01-02 b    3
2009-01-03 b    5

回答 2

如果您的csv文件包含额外的数据，则可以在导入后从DataFrame中删除列。 

import pandas as pd
from StringIO import StringIO

csv = r"""dummy,date,loc,x
bar,20090101,a,1
bar,20090102,a,3
bar,20090103,a,5
bar,20090101,b,1
bar,20090102,b,3
bar,20090103,b,5"""

df = pd.read_csv(StringIO(csv),
        index_col=["date", "loc"], 
        usecols=["dummy", "date", "loc", "x"],
        parse_dates=["date"],
        header=0,
        names=["dummy", "date", "loc", "x"])
del df['dummy']

这给了我们：

                x
date       loc
2009-01-01 a    1
2009-01-02 a    3
2009-01-03 a    5
2009-01-01 b    1
2009-01-02 b    3
2009-01-03 b    5
点击查看英文原文

If your csv file contains extra data, columns can be deleted from the DataFrame after import. 

import pandas as pd
from StringIO import StringIO

csv = r"""dummy,date,loc,x
bar,20090101,a,1
bar,20090102,a,3
bar,20090103,a,5
bar,20090101,b,1
bar,20090102,b,3
bar,20090103,b,5"""

df = pd.read_csv(StringIO(csv),
        index_col=["date", "loc"], 
        usecols=["dummy", "date", "loc", "x"],
        parse_dates=["date"],
        header=0,
        names=["dummy", "date", "loc", "x"])
del df['dummy']

Which gives us:

                x
date       loc
2009-01-01 a    1
2009-01-02 a    3
2009-01-03 a    5
2009-01-01 b    1
2009-01-02 b    3
2009-01-03 b    5

回答 3

您只需添加index_col=False参数

df1 = pd.read_csv('foo.csv',
     header=0,
     index_col=False,
     names=["dummy", "date", "loc", "x"], 
     index_col=["date", "loc"], 
     usecols=["dummy", "date", "loc", "x"],
     parse_dates=["date"])
  print df1
点击查看英文原文

You have to just add the index_col=False parameter 

df1 = pd.read_csv('foo.csv',
     header=0,
     index_col=False,
     names=["dummy", "date", "loc", "x"], 
     index_col=["date", "loc"], 
     usecols=["dummy", "date", "loc", "x"],
     parse_dates=["date"])
  print df1

回答 4

首先导入csv并使用csv.DictReader其易于处理…

点击查看英文原文

import csv first and use csv.DictReader its easy to process…

知识问答

从字符串创建Pandas DataFrame

问题：从字符串创建Pandas DataFrame

为了测试某些功能，我想DataFrame从字符串创建一个。假设我的测试数据如下：

TESTDATA="""col1;col2;col3
1;4.4;99
2;4.5;200
3;4.7;65
4;3.2;140
"""

将数据读入熊猫的最简单方法是什么DataFrame？

点击查看英文原文

In order to test some functionality I would like to create a DataFrame from a string. Let’s say my test data looks like:

TESTDATA="""col1;col2;col3
1;4.4;99
2;4.5;200
3;4.7;65
4;3.2;140
"""

What is the simplest way to read that data into a Pandas DataFrame?

回答 0

一种简单的方法是使用StringIO.StringIO（python2）或io.StringIO（python3）并将其传递给pandas.read_csv函数。例如：

import sys
if sys.version_info[0] < 3: 
    from StringIO import StringIO
else:
    from io import StringIO

import pandas as pd

TESTDATA = StringIO("""col1;col2;col3
    1;4.4;99
    2;4.5;200
    3;4.7;65
    4;3.2;140
    """)

df = pd.read_csv(TESTDATA, sep=";")
点击查看英文原文

A simple way to do this is to use StringIO.StringIO (python2) or io.StringIO (python3) and pass that to the pandas.read_csv function. E.g:

import sys
if sys.version_info[0] < 3: 
    from StringIO import StringIO
else:
    from io import StringIO

import pandas as pd

TESTDATA = StringIO("""col1;col2;col3
    1;4.4;99
    2;4.5;200
    3;4.7;65
    4;3.2;140
    """)

df = pd.read_csv(TESTDATA, sep=";")

回答 1

分割法

data = input_string
df = pd.DataFrame([x.split(';') for x in data.split('\n')])
print(df)
点击查看英文原文

Split Method

data = input_string
df = pd.DataFrame([x.split(';') for x in data.split('\n')])
print(df)

回答 2

交互式工作的快速简便解决方案是通过从剪贴板加载数据来复制和粘贴文本。

用鼠标选择字符串的内容：

在Python Shell中使用 read_clipboard()

>>> pd.read_clipboard()
  col1;col2;col3
0       1;4.4;99
1      2;4.5;200
2       3;4.7;65
3      4;3.2;140

使用适当的分隔符：

>>> pd.read_clipboard(sep=';')
   col1  col2  col3
0     1   4.4    99
1     2   4.5   200
2     3   4.7    65
3     4   3.2   140

>>> df = pd.read_clipboard(sep=';') # save to dataframe
点击查看英文原文

A quick and easy solution for interactive work is to copy-and-paste the text by loading the data from the clipboard.

Select the content of the string with your mouse:

In the Python shell use read_clipboard()

>>> pd.read_clipboard()
  col1;col2;col3
0       1;4.4;99
1      2;4.5;200
2       3;4.7;65
3      4;3.2;140

Use the appropriate separator:

>>> pd.read_clipboard(sep=';')
   col1  col2  col3
0     1   4.4    99
1     2   4.5   200
2     3   4.7    65
3     4   3.2   140

>>> df = pd.read_clipboard(sep=';') # save to dataframe

回答 3

传统的可变宽度CSV无法将数据存储为字符串变量。尤其是在.py文件内部使用时，请考虑使用定宽管道分隔数据。各种IDE和编辑器可能都有一个插件，用于将管道分隔的文本格式化为整齐的表。

使用 read_csv

将以下内容存储在实用程序模块中，例如util/pandas.py。函数的文档字符串中包含一个示例。

import io
import re

import pandas as pd


def read_psv(str_input: str, **kwargs) -> pd.DataFrame:
    """Read a Pandas object from a pipe-separated table contained within a string.

    Input example:
        | int_score | ext_score | eligible |
        |           | 701       | True     |
        | 221.3     | 0         | False    |
        |           | 576       | True     |
        | 300       | 600       | True     |

    The leading and trailing pipes are optional, but if one is present,
    so must be the other.

    `kwargs` are passed to `read_csv`. They must not include `sep`.

    In PyCharm, the "Pipe Table Formatter" plugin has a "Format" feature that can 
    be used to neatly format a table.

    Ref: https://stackoverflow.com/a/46471952/
    """

    substitutions = [
        ('^ *', ''),  # Remove leading spaces
        (' *$', ''),  # Remove trailing spaces
        (r' *\| *', '|'),  # Remove spaces between columns
    ]
    if all(line.lstrip().startswith('|') and line.rstrip().endswith('|') for line in str_input.strip().split('\n')):
        substitutions.extend([
            (r'^\|', ''),  # Remove redundant leading delimiter
            (r'\|$', ''),  # Remove redundant trailing delimiter
        ])
    for pattern, replacement in substitutions:
        str_input = re.sub(pattern, replacement, str_input, flags=re.MULTILINE)
    return pd.read_csv(io.StringIO(str_input), sep='|', **kwargs)

非工作选择

以下代码无法正常运行，因为它在左侧和右侧都添加了一个空列。

df = pd.read_csv(io.StringIO(df_str), sep=r'\s*\|\s*', engine='python')

至于read_fwf，它实际上并没有使用太多read_csv接受和使用的可选kwarg 。因此，它根本不应该用于管道分隔的数据。

点击查看英文原文

This answer applies when a string is manually entered, not when it’s read from somewhere.

A traditional variable-width CSV is unreadable for storing data as a string variable. Especially for use inside a .py file, consider fixed-width pipe-separated data instead. Various IDEs and editors may have a plugin to format pipe-separated text into a neat table.

Using read_csv

Store the following in a utility module, e.g. util/pandas.py. An example is included in the function’s docstring.

import io
import re

import pandas as pd


def read_psv(str_input: str, **kwargs) -> pd.DataFrame:
    """Read a Pandas object from a pipe-separated table contained within a string.

    Input example:
        | int_score | ext_score | eligible |
        |           | 701       | True     |
        | 221.3     | 0         | False    |
        |           | 576       | True     |
        | 300       | 600       | True     |

    The leading and trailing pipes are optional, but if one is present,
    so must be the other.

    `kwargs` are passed to `read_csv`. They must not include `sep`.

    In PyCharm, the "Pipe Table Formatter" plugin has a "Format" feature that can 
    be used to neatly format a table.

    Ref: https://stackoverflow.com/a/46471952/
    """

    substitutions = [
        ('^ *', ''),  # Remove leading spaces
        (' *$', ''),  # Remove trailing spaces
        (r' *\| *', '|'),  # Remove spaces between columns
    ]
    if all(line.lstrip().startswith('|') and line.rstrip().endswith('|') for line in str_input.strip().split('\n')):
        substitutions.extend([
            (r'^\|', ''),  # Remove redundant leading delimiter
            (r'\|$', ''),  # Remove redundant trailing delimiter
        ])
    for pattern, replacement in substitutions:
        str_input = re.sub(pattern, replacement, str_input, flags=re.MULTILINE)
    return pd.read_csv(io.StringIO(str_input), sep='|', **kwargs)

Non-working alternatives

The code below doesn’t work properly because it adds an empty column on both the left and right sides.

df = pd.read_csv(io.StringIO(df_str), sep=r'\s*\|\s*', engine='python')

As for read_fwf, it doesn’t actually use so many of the optional kwargs that read_csv accepts and uses. As such, it shouldn’t be used at all for pipe-separated data.

回答 4

最简单的方法是将其保存到临时文件，然后读取它：

import pandas as pd

CSV_FILE_NAME = 'temp_file.csv'  # Consider creating temp file, look URL below
with open(CSV_FILE_NAME, 'w') as outfile:
    outfile.write(TESTDATA)
df = pd.read_csv(CSV_FILE_NAME, sep=';')

创建临时文件的正确方法：如何在Python中创建tmp文件？

点击查看英文原文

Simplest way is to save it to temp file and then read it:

import pandas as pd

CSV_FILE_NAME = 'temp_file.csv'  # Consider creating temp file, look URL below
with open(CSV_FILE_NAME, 'w') as outfile:
    outfile.write(TESTDATA)
df = pd.read_csv(CSV_FILE_NAME, sep=';')

Right way of creating temp file: How can I create a tmp file in Python?