对于数据分析师而言,Pandas与SQL可能是大家用的比较多的两个工具,两者都可以对数据集进行深度的分析,挖掘出有价值的信息,但是二者的语法有着诸多的不同,今天小编就来总结归纳一下Pandas与SQL这两者之间在语法上到底有哪些不同。
导入数据对于Pandas而言,我们需要提前导入数据集,然后再进行进一步的分析与挖掘
import pandas as pd
airports = pd.read_csv('data/airports.csv')
airport_freq = pd.read_csv('data/airport-frequencies.csv')
runways = pd.read_csv('data/runways.csv')
基础语法
在SQL当中,我们用SELECt来查找数据,WHERe来过滤数据,DISTINCT来去重,LIMIT来限制输出结果的数量,
输出数据集
## SQL select * from airports ## Pandas airports
输出数据集的前三行数据,代码如下
## SQL select * from airports limit 3 ## Pandas airports.head(3)
对数据集进行过滤筛查
## SQL select id from airports where ident = 'KLAX' ## Pandas airports[airports.ident == 'KLAX'].id
对于筛选出来的数据进行去重
## SQL select distinct type from airport ## Pandas airports.type.unique()多个条件交集来筛选数据
多个条件的交集来筛选数据,代码如下
## SQL select * from airports where iso_region = 'US-CA' and type = 'seaplane_base' ## Pandas airports[(airports.iso_region == 'US-CA') & (airports.type == 'seaplane_base')]
或者是
## SQL select ident, name, municipality from airports where iso_region = 'US-CA' and type = 'large_airport' ## Pandas airports[(airports.iso_region == 'US-CA') & (airports.type == 'large_airport')][['ident', 'name', 'municipality']]排序
在Pandas当中默认是对数据进行升序排序,要是我们希望对数据进行降序排序,需要设定ascending参数
## SQL
select * from airport_freq
where airport_ident = 'KLAX'
order by type
## Pandas
airport_freq[airport_freq.airport_ident == 'KLAX']
.sort_values('type')
又或者是
## SQL
select * from airport_freq
where airport_ident = 'KLAX'
order by type desc
## Pandas
airport_freq[airport_freq.airport_ident == 'KLAX']
.sort_values('type', ascending=False)
筛选出列表当中的数据
要是我们需要筛选出来的数据在一个列表当中,这里就需要用到isin()方法,代码如下
## SQL
select * from airports
where type in ('heliport', 'balloonport')
## Pandas
airports[airports.type.isin(['heliport', 'balloonport'])]
又或者是
## SQL
select * from airports
where type not in ('heliport', 'balloonport')
## Pandas
airports[~airports.type.isin(['heliport', 'balloonport'])]
删除数据
在Pandas当中删除数据用的是drop()方法,代码如下
## SQL delete from dataframe where col_name = 'MISC' ## Pandas df = df[df.type != 'MISC'] df.drop(df[df.type == 'MISC'].index)更新数据
在SQL当中更新数据使用的是update和set方法,代码如下
### SQL update airports set home_link = '......' where ident == 'KLAX' ### Pandas airports.loc[airports['ident'] == 'KLAX', 'home_link'] = '......'调用统计函数
对于给定的数据集,如下图所示
runways.head()
output
我们调用min()、max()、mean()以及median()函数作用于length_ft这一列上面,代码如下
## SQL
select max(length_ft), min(length_ft),
avg(length_ft), median(length_ft) from runways
## Pandas
runways.agg({'length_ft': ['min', 'max', 'mean', 'median']})
合并两表格
在Pandas当中合并表格用的是pd.concat()方法,在SQL当中则是UNIOn ALL,代码如下
## SQL select name, municipality from airports where ident = 'KLAX' union all select name, municipality from airports where ident = 'KLGB' ## Pandas pd.concat([airports[airports.ident == 'KLAX'][['name', 'municipality']], airports[airports.ident == 'KLGB'][['name', 'municipality']]])分组
顾名思义也就是groupby()方法,代码如下
## SQL select iso_country, type, count(*) from airports group by iso_country, type order by iso_country, type ## Pandas airports.groupby(['iso_country', 'type']).size()分组之后再做筛选
在Pandas当中是在进行了groupby()之后调用filter()方法,而在SQL当中则是调用HAVINg方法,代码如下
## SQL
select type, count(*) from airports
where iso_country = 'US'
group by type
having count(*) > 1000
order by count(*) desc
## Pandas
airports[airports.iso_country == 'US']
.groupby('type')
.filter(lambda g: len(g) > 1000)
.groupby('type')
.size()
.sort_values(ascending=False)
TOP N records
代码如下
## SQL select 列名 from 表名 order by size desc limit 10 ## Pandas 表名.nlargest(10, columns='列名')
在公众号后台回复【20220317】获取源数据
NO.1
往期推荐
Historical articles
Pandas多层级索引的数据分析案例,超干货的!
用Python制作数据大屏,超简单
厉害了,用Python绘制动态可视化图表,并保存成gif格式
20个Pandas数据实战案例,干货多多
分享、收藏、点赞、在看安排一下?



