【Python_017】sql语法在pandas中的实现

平时SQL用惯了，用python处理起数据来有些不大顺手，于是整理了一下sql基本常用语法在pandas中的实现。

本篇文章均已以下数据为例演示：

distinct

• SQL

  select distinct city from table
  

• Python

  data['city'].drop_duplicates()
  

  

Group by

• SQL

  select city, sum(sales) as Q2_sales
  from table
  group by city	
  

• Python

  data.groupby('city')['sales'].sum().reset_index().rename(columns={'sales':'Q2_sales'})
  

  

Where

定位到某行或多个字段

• SQL

  select *
  from table
  where month = 202006
  

• Python

  data[data['month']== 202006 ]
  

  

定位到某个具体的值

• SQL

  select sales
  from table
  where city='上海' and month = 202005
  

• Python

  data.[(data['city']=='上海') & (data['month']==202005)]['sales'].values[0]
  

  

where条件中加in

• SQL

  select *
  from table
  where month in (202005, 202006)
  

• Python

  data[data['month'].isin([202005,202006])]
  ##若需要not in 则在条件前加-就好
  #data[-data['month'].isin([202005,202006])]
  

  

join

之前整理过一篇 pandas中merge用法, pandas中的merge便和sql中的join是相同效果，在此不多赘述，戳链接就好

Order by

• SQL

  select *
  from table 
  order by month asc, sales desc
  

• Python

  data.sort_values(['month','sales'],ascending=[True,False])
  

  

Top

• SQL

  select top 3 *
  from table
  

• Python

  data.nsmallest(4,columns=['sales'])
  data.nlargest(4,columns=['sales'])
  

  当然也可以sort_values之后再去head() / tail()