Mastering Python for Data Science

相似文件 换一批

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

本帖隐藏的内容

Mastering Python for Data Science_PACKT_2015.pdf (6.03 MB)

2015-10-2 00:44:18 上传

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏3 回帖

关键词：Data Science Mastering Science python Master

1. Creating multiple plots
   
2. 
   
3. One very useful feature of matplotlib is that it makes it easy to plot multiple plots, which can be compared to each other:
   
4. 
   
5. >>> p1 = np.arange(0.0, 30.0, 0.1)
   
6. 
   
7. >>> plt.subplot(211)
   
8. >>> plt.plot(p1, np.sin(p1)/p1, 'b--')
   
9. 
   
10. >>> plt.subplot(212)
    
11. >>> plt.plot(p1, np.cos(p1), 'r--')
    
12. >>> plt.show()

复制代码

1. Playing with text
   
2. 
   
3. Adding text to your chart can be done by using a simple matplotlib function. You only have to use the text() command to add it to the chart:
   
4. 
   
5. >>> # Playing with text
   
6. >>> n = np.random.random_sample((5,))
   
7. 
   
8. >>> plt.bar(np.arange(len(n)), n)
   
9. >>> plt.xlabel('Indices')
   
10. >>> plt.ylabel('Value')
    
11. >>> plt.text(1, .7, r'$\mu=' + str(np.round(np.mean(n), 2)) + '
    
12. )
    
13. 
    
14. >>> plt.show()
    
15. In the preceding code, the text() com

复制代码

1. Styling your plots
   
2. 
   
3. The style package within the matplotlib library makes it easier to change the style of the plots that are being plotted. It is very easy to change to the famous ggplot style of the R language or use the Nate Silver's website http://fivethirtyeight.com/ for fivethirtyeight style. The following example shows the plotting of a simple line chart with the ggplot style:
   
4. 
   
5. >>> plt.style.use('ggplot')
   
6. >>> plt.plot([1, 2, 3, 4], [1, 4, 9, 16])
   
7. >>> plt.show()

复制代码

1. Area plots
   
2. 
   
3. An area plot is useful for comparing the values of different factors across a range. The area plot can be stacked in nature, where the areas of the different factors are stacked on top of each other. The following code gives an example of a stacked area plot:
   
4. 
   
5. >>> df = pd.DataFrame(np.random.rand(10, 4), columns=['p', 'q', 'r', 's'])
   
6. 
   
7. >>> df.plot(kind='area');

复制代码

1. Heatmaps
   
2. 
   
3. A heatmap is a graphical representation where individual values of a matrix are represented as colors. A heatmap is very useful in visualizing the concentration of values between two dimensions of a matrix. This helps in finding patterns and gives a perspective of depth.
   
4. 
   
5. Let's start off by creating a basic heatmap between two dimensions. We'll create a 10 x 6 matrix of random values and visualize it as a heatmap:
   
6. 
   
7. >>> # Generate Data
   
8. >>> data = np.random.rand(10,6)
   
9. >>> rows = list('ZYXWVUTSRQ')  #Ylabel
   
10. >>> columns = list('ABCDEF')  #Xlabel
    
11. 
    
12. >>> #Basic Heat Map plot
    
13. >>> plt.pcolor(data)
    
14. >>> plt.show()

复制代码

1. Scatter plots with histograms
   
2. 
   
3. We can combine a simple scatter plot with histograms for each axis. These kinds of plots help us see the distribution of the values of each axis.
   
4. 
   
5. Let's generate some randomly distributed data for the two axes:
   
6. 
   
7. >>> from matplotlib.ticker import NullFormatter
   
8. >>> # the random data
   
9. >>> x = np.random.randn(1000)
   
10. >>> y = np.random.randn(1000)
    
11. A NullFormatter object is created, which will be used for eliminating the x and y labels of the histograms:
    
12. 
    
13. >>> nullfmt   = NullFormatter()         # no labels
    
14. The following code defines the size, height, and width of the scatter and histogram plots:
    
15. 
    
16. >>> # definitions for the axes
    
17. >>> left, width = 0.1, 0.65
    
18. >>> bottom, height = 0.1, 0.65
    
19. >>> bottom_h = left_h = left+width+0.02

复制代码

1. A scatter plot matrix
   
2. 
   
3. A scatter plot matrix can be formed for a collection of variables where each of the variables will be plotted against each other. The following code generates a DataFrame df, which consists of four columns with normally distributed random values and column names named from a to d:
   
4. 
   
5. >>> df = pd.DataFrame(np.random.randn(1000, 4), columns=['a', 'b', 'c', 'd'])
   
6. 
   
7. >>> spm = pd.tools.plotting.scatter_matrix(df, alpha=0.2, figsize=(6, 6), diagonal='hist')

复制代码

xiexie dada

1. Bubble charts
   
2. 
   
3. A bubble chart is basically a scatter plot with an additional dimension. The additional dimension helps in setting the size of the bubble, which means that the greater the size of the bubble, the larger the value that represents the bubble. This kind of a chart helps in analyzing the data of three dimensions.
   
4. 
   
5. The following code creates a sample data of three variables and this data is then fed to the plot() method where its kind is mentioned as a scatter and s is the size of the bubble:
   
6. 
   
7. >>> plt.style.use('ggplot')
   
8. >>> df = pd.DataFrame(np.random.rand(50, 3), columns=['a', 'b', 'c'])
   
9. >>> df.plot(kind='scatter', x='a', y='b', s=df['c']*400);
   
10. After the preceding code is executed we'll get the following output:

复制代码