|
Python用于數(shù)據(jù)分析好不好，因人而異。但是，excel能做的Python也能做。比如聚合，透視表。我認為更簡單一下。如果大家在工作中用得到數(shù)據(jù)庫，用得到Python，可以學習一下數(shù)據(jù)透視表。 import pymysqlimport pandas as pdimport matplotlib.pyplot as plt
# 建立連接conn = pymysql.connect('localhost','username','password','database')
# 讀取SQL為dfsql = 'select * from table'df = pd.read_sql(sql,con=conn)
# 查看基本屬性print(df.index)print(df.columns)print(df.info())
# 修改時間格式df['stat_month'] = pd.to_datetime(df['stat_month'],format='%Y%m')
# 設(shè)置索引df.set_index('stat_month',inplace = True)print(df.head())
# 只看一個月份的df_grp1601= df['20160101'].groupby(['brand','areaname'],as_index = False)['profit'].sum()print(df_grp1601)
df_grpbrand = df_grp1601.groupby('brand').sum()print(df_grpbrand[df_grpbrand>200000].plot(kind = 'bar'))
pd.pivot_table(df['20160101':'20161201'],values = 'profit',\ index = ['stat_month','brand'],\ columns ='areaname',aggfunc='sum' )
祝大家早日富可敵國！
記得 點 在看 哦
|
