《Python数据分析技术栈》第06章使用 Pandas 准备数据 02 pandas的组件（Building blocks of Pandas）

02 pandas的组件（Building blocks of Pandas）

《Python数据分析技术栈》第06章使用 Pandas 准备数据 02 pandas的组件（Building blocks of Pandas）

The Series and DataFrame objects are the underlying data structures in Pandas. In a nutshell, a Series is like a column (has only one dimension), and a DataFrame (has two dimensions ) is like a table or a spreadsheet with rows and columns. Each value stored in a Series or a DataFrame has a label or an index attached to it, which speeds up retrieval and access to data. In this section, we learn how to create a Series and DataFrame, and the functions used for manipulating these objects.

系列（Series）和数据帧（DataFrame）对象是 Pandas 的底层数据结构。简而言之，Series 就像一列（只有一个维度），而 DataFrame（有两个维度）就像一个有行和列的表格或电子表格。存储在系列或 DataFrame 中的每个值都附有标签或索引，这可以加快数据检索和访问速度。在本节中，我们将学习如何创建系列和数据帧，以及用于操作这些对象的函数。

Creating a Series object

The Series is a one-dimensional object, with a set of values and their associated indexes. Table 6-1 lists the different ways of creating a series.

系列是一个一维对象，包含一组值及其相关索引。表 6-1 列出了创建系列的不同方法。

Using a scalar value：

使用标量值：

pd.Series(2)
#Creating a simple series with just one value. Here, 0 is the index label, and 2 is the value the Series object contains.

Using a list：

pd.Series([2]*5)
#Creating a series by enclosing a single value (2) in a list and replicating it 5 times. 0,1,2,3,4 are the autogenerated index labels.

Using characters in a string:

pd.Series(list('hello'))
#Creating a series by using each character in the string "hello" as a separate value in the Series.

Using a dictionary:

pd.Series({1:'India',2:'Japan',3:'Singapore'})
#The key/value pairs correspond to the index labels and values in the Series object.

Using a range:

pd.Series(np.arange(1,5))
#Using the NumPy arrange function to create a series from a range of 4 numbers (1-4), ensure that the NumPy library is also imported

Using random numbers:

pd.Series(np.random.normal(size=4))
#Creating a set of 4 random numbers using the np.random.normal function

Creating a series with customized index labels:

pd.Series([2,0,1,6],index=['a','b','c','d'])
#The list [2,0,1,6] specifies the values in the series, and the list for the index['a','b','c','d'] specifies the index labels

To summarize, you can create a Series object from a single (scalar) value, list, dictionary, a set of random numbers, or a range of numbers. The pd.Series function creates a Series object (note that the letter “S” in “Series” is in uppercase; pd.series will not work). Use the index parameter if you want to customize the index.

简而言之，您可以通过单个（标量）值、列表、字典、随机数集或数字范围创建 Series 对象。函数 pd.Series 将创建一个 Series 对象（注意，"Series "中的字母 "S "是大写字母；pd.series 将不起作用）。如果要自定义索引，请使用索引参数。