巧用Superset大数据分析平台搞定各类图表

旧城等待， 2022-06-07 00:45 193阅读 0赞

# http://m.blog.csdn.net/qq273681448/article/details/75050513    #

# 前言 #

其实大数据图表展示的这类平台有很多，Superset是其中之一，最近有个需求对各类图表展示的开发较多，索性将工作量交给这个平台。

# 介绍 #

Superset的中文翻译是快船，而Superset其实是一个自助式数据分析工具，它的主要目标是简化我们的数据探索分析操作，它的强大之处在于整个过程一气呵成，几乎不用片刻的等待。  
![ed500a2f6ee6deef8ed46ca9c2588989][]

# 部署 #

## docker方式（推荐） ##

docker pull amancevice/caravel
    docker run --name caravel -d -p 8088:8088 amancevice/caravel
    docker exec -it caravel demo
    #系统会提示设置管理员
        
         
          1 
          2 
          3 
          4

具体效果如下

[root@localhost ~]# docker exec -it ac94ec4d9469 demo
    Username [admin]: admin
    User first name [admin]: admin
    User last name [user]: a
    Email [admin@fab.org]: a
    Password: 
    Repeat for confirmation:
    ******此处省略几百行
        
         
          1 
          2 
          3 
          4 
          5 
          6 
          7 
          8

## centos系统 ##

yum makecache
    sudo easy_install -i http://pypi.douban.com/simple/ pip
    sudo easy_install pip
    sudo yum install gcc libffi-devel python-devel python-pip python-wheel openssl-devel libsasl2-devel openldap-devel
        
         
          1 
          2 
          3 
          4

#安装Superset 
    pip install superset 
    
    #创建管理员用户名和密码 
    fabmanager create-admin --app superset 
    
    #初始化Superset 
    superset db upgrade 
    
    #装载初始化样例数据 
    superset load_examples 
    
    #创建默认角色和权限 
    superset init 
    
    #启动Superset 
    superset runserver 
        
         
          1 
          2 
          3 
          4 
          5 
          6 
          7 
          8 
          9 
          10 
          11 
          12 
          13 
          14 
          15 
          16 
          17

# 数据库依赖 #

<table style="border-collapse:collapse; border-spacing:0px; max-width:100%; background-color:transparent; width:1240px; border:0.05rem solid rgb(238,238,238); word-break:break-all"> 
 <thead style=""> 
  <tr style=""> 
   <th style="text-align:left; padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> 数据库</th> 
   <th align="left" style="text-align:left; padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> 命令</th> 
   <th align="left" style="text-align:left; padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> 链接方式</th> 
  </tr> 
 </thead> 
 <tbody style=""> 
  <tr style=""> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> MySQL</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> pip install mysqlclient</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> mysql://</td> 
  </tr> 
  <tr style=""> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> Postgres</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> pip install psycopg2</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> postgresql+psycopg2://</td> 
  </tr> 
  <tr style=""> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> Presto</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> pip install pyhive</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> presto://</td> 
  </tr> 
  <tr style=""> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> Oracle</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> pip install cx_Oracle</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> oracle://</td> 
  </tr> 
  <tr style=""> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> sqlite</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> &nbsp;</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> sqlite://</td> 
  </tr> 
  <tr style=""> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> Redshift</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> pip install sqlalchemy-redshift</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> redshift+psycopg2://</td> 
  </tr> 
  <tr style=""> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> MSSQL</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> pip install pymssql</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> mssql://</td> 
  </tr> 
  <tr style=""> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> Impala</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> pip install impyla</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> impala://</td> 
  </tr> 
  <tr style=""> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> SparkSQL</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> pip install pyhive</td> 
   <td align="left" style="padding:8px; line-height:20px; vertical-align:top; border:0.05rem solid rgb(238,238,238)"> jdbc+hive://</td> 
  </tr> 
 </tbody> 
</table>

# 改造（docker环境） #

因为默认是有权限控制的，即使分享出去的图表也需要登录用户才能看到，那如何做到不登陆用户也能查看图表呢，于是改造之路就开始了。

## 方法一：改源码 ##

刚开始没找到方法，不得不从从源码入手解开权限，  
[https://github.com/apache/incubator-superset/blob/master/superset/views/core.py][https_github.com_apache_incubator-superset_blob_master_superset_views_core.py]这个文件  
/superset/views/core.py文件922行删掉 @has\_access\_api注解

@log_this
        @has_access_api
        @expose("/explore_json/<datasource_type>/<datasource_id>/")
        def explore_json(self, datasource_type, datasource_id):
            try:
                viz_obj = self.get_viz(
                    datasource_type=datasource_type,
        
         
          1 
          2 
          3 
          4 
          5 
          6 
          7

1006行删掉 @has\_access\_api注解

@log_this
        @has_access
        @expose("/explore/<datasource_type>/<datasource_id>/")
        def explore(self, datasource_type, datasource_id):
            form_data = self.get_form_data()
        
         
          1 
          2 
          3 
          4 
          5

在1017行加入

viz_obj.datasource.("all_datasource_access")
        
         
          1

增加库的访问权限，可能还有些地方要改，该方法改了源码，侵入性强，不推荐。

## 方法二：改配置（推荐） ##

/superset/config.py第127行

# ---------------------------------------------------
    # Roles config
    # ---------------------------------------------------
    # Grant public role the same set of permissions as for the GAMMA role.
    # This is useful if one wants to enable anonymous users to view
    # dashboards. Explicit grant on specific datasets is still required.
    PUBLIC_ROLE_LIKE_GAMMA = False
    
        
         
          1 
          2 
          3 
          4 
          5 
          6 
          7 
          8

将`PUBLIC_ROLE_LIKE_GAMMA`改为True，  
注释意思  
授予公共角色与GAMMA角色相同的权限集。  
如果想让匿名用户查看，可以设置这里  
在仪表盘对特定数据集的授权显示，也在这里设置。

加入数据库权限这里加入所有数据库权限

![c39d8ecc5ca6a3db88c85358d90b13fb][]

Public  [can explore on Superset, all database access on all_database_access, can explore json on Superset]
        
         
          1

其中  
can explore on Superset为导出图表  
can explore json on Superset为导出图表json  
all database access on all\_database\_access访问所有数据库权限，也可以设置单个  
![619458b2d8a801a1548ccaca1cf2935d][]

# iframe嵌入 #

![0b610cc7f41eb60b5882d4eb1cb6eff4][]  
选择某个切片，点右上角

将iframe放入任意html内即可打开。

# 更多样式 #

![18fa2606c93042069e653d9ad3827e06][]  
![c021af152ffda53f51e48dddf0d3a4d2][]

![863df8ee6a777dcf7277117899915b56][]

# 总结 #

借助superset可以帮助我们可视化进行数据库分析，支持导出导入sql语句，支持各种数据库，在图表展现方面是个不错的框架。

觉得好的话，记得关注我哦！  
掘金：  
[https://juejin.im/user/57cd55218ac247006459c40c][https_juejin.im_user_57cd55218ac247006459c40c]  
Github：  
[https://github.com/qq273681448][https_github.com_qq273681448]

[ed500a2f6ee6deef8ed46ca9c2588989]: https://user-gold-cdn.xitu.io/2017/7/13/ed500a2f6ee6deef8ed46ca9c2588989
[https_github.com_apache_incubator-superset_blob_master_superset_views_core.py]: https://github.com/apache/incubator-superset/blob/master/superset/views/core.py
[c39d8ecc5ca6a3db88c85358d90b13fb]: https://user-gold-cdn.xitu.io/2017/7/13/c39d8ecc5ca6a3db88c85358d90b13fb
[619458b2d8a801a1548ccaca1cf2935d]: https://user-gold-cdn.xitu.io/2017/7/13/619458b2d8a801a1548ccaca1cf2935d
[0b610cc7f41eb60b5882d4eb1cb6eff4]: https://user-gold-cdn.xitu.io/2017/7/13/0b610cc7f41eb60b5882d4eb1cb6eff4
[18fa2606c93042069e653d9ad3827e06]: https://user-gold-cdn.xitu.io/2017/7/13/18fa2606c93042069e653d9ad3827e06
[c021af152ffda53f51e48dddf0d3a4d2]: https://user-gold-cdn.xitu.io/2017/7/13/c021af152ffda53f51e48dddf0d3a4d2
[863df8ee6a777dcf7277117899915b56]: https://user-gold-cdn.xitu.io/2017/7/13/863df8ee6a777dcf7277117899915b56
[https_juejin.im_user_57cd55218ac247006459c40c]: https://juejin.im/user/57cd55218ac247006459c40c
[https_github.com_qq273681448]: https://github.com/qq273681448