分箱可以將連續(xù)變量離散化，減小異常值對(duì)模型的影響

數(shù)據(jù)準(zhǔn)備

Age = [0,10,20,25,31,35,40,62,90]

pd.qcut() 使每一份的元素個(gè)數(shù)相同

#將Age分為三個(gè)箱子，每個(gè)箱子有3個(gè)元素
pd.qcut(data['Age'],3,labels=['Teen',‘Middle-age’,'Elder'])

<<[Teen, Teen, Teen, Middle-age, Middle-age, Middle-age, Elder, Elder, Elder]

pd.cut 使每一份的寬度相同

#將Age分為三個(gè)箱子，箱子范圍分別是0-30，30-60，60-90
pd.cut(Age,3,labels=['Teen',‘Middle-age’,'Elder'])

<<<[Teen, Teen, Teen, Teen, Middle-age, Middle-age, Middle-age, Elder, Elder]

給Age指定區(qū)間和標(biāo)簽

pd.cut(ages, [0,5,20,30,50,100], labels=[u"嬰兒",u"青年",u"中年",u"壯年",u"老年"])