pandas数据分析之绘图方法详解

左手的ㄟ右手 2023-02-20 07:56 7阅读 0赞

之前介绍过[matplotlib plot方法的详细用法][matplotlib plot]，今天来介绍一下pandas中绘图方法，pandas中封装的绘图方法本质上还是调用了matplotlib的绘图方法。

*  基础绘图方法plot，pandas中Series和DataFrame的plot方法是包装自matplotlib的plot方法

# Series绘图
    ts = pd.Series(
        np.random.randn(1000),
        index=pd.date_range('1/1/2000', periods=1000)
    )
    # cumsum累加求和，并保留累加的中间结果
    ts = ts.cumsum(axis=0)
    ts.plot()

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70][]

*  对DataFrame调用plot函数，会将各个属性的名称设置为labels

df = pd.DataFrame(
        np.random.randn(1000, 4),
        index=ts.index,
        columns=["A", "B", "C", "D"]
    )
    df = df.cumsum(axis=0)
    df.plot()

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 1][]

*  DataFrame调用plot函数时，可以通过指定x和y参数来绘制两个属性之间的关系

df3 = pd.DataFrame(
        np.random.randn(100, 2),
        columns=["B", "C"]
    )
    df3["A"] = pd.Series(list(range(len(df3))))
    """
        参数解释：详见matplotlib plot函数参数解释
        x：横轴数据
        y：纵轴数据
        color绘图颜色
        linestyle:线条样式
        linewidth:线条宽度
        figsize:绘图大小
        legend:是否显示图例
        logy:是否对y轴数据进行log变化，控制数值范围
    """
    df3.plot(x="A", y="B", color="b", linestyle="-", linewidth=2, figsize=(10,10), legend=True, logy=True)

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 2][]

绘制线形图以外的其他图像时，可以通过plot函数的kind参数控制图像样式：

*  ‘bar’ or ‘barh’ for bar plots -> 条形图（竖直或水平）
 *  ‘hist’ for histogram -> 柱状图
 *  ‘box’ for boxplot -> 箱型图
 *  ‘kde’ or ‘density’ for density plots -> 密度图
 *  ‘area’ for area plots -> 面积图
 *  ‘scatter’ for scatter plots -> 散点图
 *  ‘hexbin’ for hexagonal bin plots -> 六角本图
 *  ‘pie’ for pie plots -> 饼图

1）、kind="bar"条形图(竖直)

df.iloc[5].plot(kind="bar")
    print(df.iloc[5])

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 3][]

2）、kind="barh"条形图（水平）

df.iloc[5].plot(kind="barh")
    print(df.iloc[5])

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 4][]

3）、除了使用plot的kind参数指定绘图样式，也可以直接使用plot.的形式指定绘图样式

# 绘制条形图
    df.iloc[5].plot.bar()

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 5][]

*  绘制条形图DataFrame.plot.bar()

df.iloc[5].plot.bar()
    plt.axhline(0, color='b');

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 6][]

df2 = pd.DataFrame(np.random.rand(10, 4), columns=['a', 'b', 'c', 'd'])
    df2.plot.bar()

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 7][]

# 将条形图进行堆叠
    df2.plot.bar(stacked=True)

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 8][]

# 绘制水平条形图
    df2.plot.barh(stacked=True)

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 9][]

*  绘制柱状图DataFrame.plot.hist()

df = pd.DataFrame(
        {
            "a": np.random.randn(1000) + 1,
            "b": np.random.randn(1000),
            "c": np.random.randn(1000) - 1
        },
        columns=["a", "b", "c"]
    )
    # alpha参数控制图像的透明度
    df.plot.hist(alpha=0.5)

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 10][]

# 柱状图堆叠，并指定柱状图的分箱数
    #  bins参数控制统计区间的大小
    df.plot.hist(stacked=True, bins=20)

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 11][]

# 也可以使用matplotlib库hist方法提供的关键字参数
    df.plot.hist(orientation="horizontal", cumulative=True)

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 12][]

# 除了通过df.plot.hist方式进行调用，DataFrame也有hist方法
    # DataFrame的hist方法使用多个子图分别绘制DataFrame的column
    df.hist(bins=20)

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 13][]

s = pd.Series([1,3,4,6,7,9])
    # diff函数用来判断属性的每个值与指定数值的差距，默认的指定数值为前一行的数值内容
    s.diff().hist()

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 14][]

# DataFrame的hist方法可以使用by参数来定义分组
    data = pd.Series(np.random.randn(1000))
    # np.random.randint(0, 4, 1000)随机生成1000个0-3之间的数字，然后将这些数字进行分组，并将与之对应的data数据绘制直方图
    data.hist(by=np.random.randint(0, 4, 1000), figsize=(6, 4))
    pd.Series(np.random.randint(0, 4, 1000)).value_counts()

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 15][]

*  绘制箱型图DataFrame.plot.box()

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 16][]

箱型图简介：   
（1）计算上四分位数（Q3），中位数，下四分位数（Q1）   
（2）计算上四分位数和下四分位数之间的差值，即四分位数差（IQR，interquartile range）Q3-Q1   
（3）绘制箱线图的上下范围，上限为上四分位数，下限为下四分位数。在箱子内部中位数的位置绘制横线。   
（4）大于 上四分位数 1.5倍四分位数差 的值，或者小于下四分位数1.5倍四分位数差的值，划为异常值（outliers）。   
（5）异常值之外，最靠近上边缘和下边缘的两个值处，画横线，作为箱线图的触须。   
（6）极端异常值，即超出四分位数差3倍距离的异常值，用实心点表示；较为温和的异常值，即处于1.5倍-3倍四分位数差之间的异常值，用空心点表示。   
（7）为箱线图添加名称，数轴等

df = pd.DataFrame(np.random.randn(10, 5), columns=["A", "B", "C", "D", "E"])
    # 绘制所有数值属性的箱线图
    df.plot.box()

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 17][]

# 绘制单独一个属性的箱线图
    df["A"].plot.box()

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 18][]

df.boxplot()

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 19][]

# 设置箱线图的颜色
    color = {'boxes': 'DarkGreen', 'whiskers': 'DarkOrange', 'medians': 'DarkBlue', 'caps': 'Gray'}
    # sym指定异常值的样式
    df.plot.box(color=color, sym='r+')

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 20][]

*  绘制面积图DataFrame.plot.area()

面积图的意义：堆积面积图是一种特殊的面积图，可以用来比较在一个区间内的多个变量。堆叠面积图和面积图的区别在于，堆叠面积图每个数据系列的起点都是基于前一个数据系列绘制的（堆叠面积图要求一列数据要么全为正，要么全为负），也就是每度量一行就要填满行与行之间的区域。如果有多个数据系列，并想分析每个类别的部分到整体的关系，并展现部分量对于总量的贡献时，使用堆积面积图是非常合适的选择。

df = pd.DataFrame(np.random.rand(10, 4), columns=["A", "B", "C", "D"])
    df["B"] = df["B"] * -1.0
    # 绘制的面积图默认是堆叠的形式，堆叠面积图要求同一个特征的数值要么全为正，要么全为负
    df.plot.area()
    df.loc[0].sum() - df.loc[0, "B"], df.loc[0, "B"], df.loc[9].sum() - df.loc[9, "B"], df.loc[9, "B"]

(1.2377821011551786, -0.3169154766752843, 1.9799948518694093, -0.7935035073522937)

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 21][]

df.loc[0:5, "B"] = df.loc[0:5, "B"] * -1.0
    df.plot.area(stacked=False)

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 22][]

*  绘制散点图DataFrame.plot.scatter()

散点图要求属性是数值型的。

df = pd.DataFrame(np.random.rand(50, 4), columns=['a', 'b', 'c', 'd'])
    df.plot.scatter(x="a", y="b")

# 在一个axes中绘制多个散点图
    # 获取画布的句柄，指定绘制的颜色，标记样式
    ax = df.plot.scatter(x="a", y="b", color="r", marker=".", figsize=(10, 8), s=50)
    # 指定画布，颜色，标记样式，用来区分不同的数据
    df.plot.scatter(x="c", y="d", ax=ax, color="b", marker="^")
    df.plot.scatter(x="a", y="c", ax=ax, color="k", marker="<")

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 23][]

# 使用关键字c来为每个数据点设置不同的颜色，通常用来显示两个变量（x,y）的共同作用对第三个变量c的影响
    df.plot.scatter(x='a', y='b', c='c', s=50)

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 24][]

*  不同绘图方法的缺失值处理方法

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 25][]

*  绘制散点图矩阵pandas.plotting.scatter\_matrix

from pandas.plotting import scatter_matrix
    
    # 使用scatter_matrix绘制所有特征组合之间的散点图，并指定对角线上的显示样式密度kde或柱状图hist
    df = pd.DataFrame(np.random.randn(1000, 4), columns=['a', 'b', 'c', 'd'])
    scatter_matrix(df, alpha=0.2, figsize=(8, 8), diagonal="hist")

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 26][]

*  绘制密度图DataFrame.plot.kde()

df.plot.kde(legend=True)

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 27][]

参考：[pandas官网][pandas] [pandas CookBook][]

道阻且长，加油吧，少年！

[matplotlib plot]: https://blog.csdn.net/cxx654/article/details/106363371
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70]: https://img-blog.csdnimg.cn/20200625121223945.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 1]: https://img-blog.csdnimg.cn/20200625121315597.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 2]: https://img-blog.csdnimg.cn/20200625121430133.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 3]: https://img-blog.csdnimg.cn/20200625121624338.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 4]: https://img-blog.csdnimg.cn/20200625121704468.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 5]: https://img-blog.csdnimg.cn/20200625121737634.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 6]: https://img-blog.csdnimg.cn/20200625121900204.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 7]: https://img-blog.csdnimg.cn/20200625121938363.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 8]: https://img-blog.csdnimg.cn/20200625122011491.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 9]: https://img-blog.csdnimg.cn/20200625122038601.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 10]: https://img-blog.csdnimg.cn/20200625122106446.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 11]: https://img-blog.csdnimg.cn/20200625122133765.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 12]: https://img-blog.csdnimg.cn/20200625122207647.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 13]: https://img-blog.csdnimg.cn/20200625122234498.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 14]: https://img-blog.csdnimg.cn/20200625122258954.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 15]: https://img-blog.csdnimg.cn/20200625122326477.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 16]: https://img-blog.csdnimg.cn/20200625122408808.jpeg?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 17]: https://img-blog.csdnimg.cn/20200625122558310.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 18]: https://img-blog.csdnimg.cn/20200625122721351.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 19]: https://img-blog.csdnimg.cn/2020062512274267.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 20]: https://img-blog.csdnimg.cn/20200625122807860.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 21]: https://img-blog.csdnimg.cn/20200625122959203.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 22]: https://img-blog.csdnimg.cn/20200625123033384.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 23]: https://img-blog.csdnimg.cn/20200625123201323.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 24]: https://img-blog.csdnimg.cn/20200625123333315.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 25]: https://img-blog.csdnimg.cn/20200625123408246.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 26]: https://img-blog.csdnimg.cn/20200625123556241.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA_size_16_color_FFFFFF_t_70 27]: https://img-blog.csdnimg.cn/20200625123716130.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2N4eDY1NA==,size_16,color_FFFFFF,t_70
[pandas]: https://pandas.pydata.org/docs/user_guide/visualization.html
[pandas CookBook]: https://pandas.pydata.org/docs/user_guide/cookbook.html#cookbook-plotting