这组数据如何取

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

这样一组数据

var1	market	date	corp
2005-04-08,五		20050408
2005-04-11,一	0.78	20050411	3.77
2005-04-12,二	1.7	20050412	0.66
2005-04-13,三	2.27	20050413	0.13
2005-04-14,四	1.39	20050414	0.39
2005-04-15,五	1.31	20050415	3.15
2005-04-18,一	1.06	20050418	0.25
2005-04-19,二	0.22	20050419	1.84
2005-04-20,三	1.55	20050420	0.19
2005-04-21,四	0.72	20050421	3.4
2005-04-22,五	0.52	20050422	0.5
2005-04-25,一	0.96	20050425	2.99
2005-04-26,二	0.75	20050426	1.51
2005-04-27,三	1.12	20050427	1.84
2005-04-28,四	1.67	20050428	1.27
2005-04-29,五	1.03	20050429	6.11
2005-05-09,一	2.49	20050509	9.7
2005-05-10,二	0.43	20050510	2.81
2005-05-11,三	1.23	20050511	6.17
2005-05-12,四	1.78	20050512	3.06
2005-05-13,五	0.19	20050513	0

比如我要去20050510这一天前的5天和后3天的数据怎么取？

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：market marke Mark date Corp 数据

clear
use C:\Users\Administrator\Desktop\sz\sz\zhongji.dta
drop in 1
drop in 1
gen date=substr(var1,1,10)
destring date var2,ignore("-") replace
rename var2 corp
sort date
save zhongji,replace
clear
use C:\Users\Administrator\Desktop\sz\sz\market.dta
drop in 1
drop in 1
gen date=substr(var1,1,10)
destring date var2,ignore("-") replace
rename var2 market
sort date
save market,replace
merge date using zhongji
drop _merge
drop if corp==.
drop if market==.
gen corp1=corp if date<20070803
gen market1=market if date<20070803
sort date
drop if corp1==.
drop if market==.
save cm1,replace
save C:\Users\Administrator\Desktop\sz\cm1,replace
clear
use C:\Users\Administrator\Desktop\sz\sz\zhongji.dta
drop in 1
drop in 1
gen date=substr(var1,1,10)
destring,ignore("-") replace
rename var2 corp
sort date
save zhongji,replace
clear
use C:\Users\Administrator\Desktop\sz\sz\market.dta
drop in 1
drop in 1
gen date=substr(var1,1,10)
destring,ignore("-") replace
rename var2 market
sort date
save market,replace
merge date using zhongji
drop _merge
drop if corp==.
drop if market==.
gen corp2=corp if date>=20070803
gen market2=market if date>=20070803
sort date
drop if corp2==.
drop if market2==.
save cm2,replace
clear
use cm1
sort date
gen corp3=corp1 in -105/-6
gen market3=market1 in -105/-6
gen date1=date in -105/-6
sort date1
drop if corp3==.
drop date corp1 market1
save cm3,replace
clear
use C:\Users\Administrator\Desktop\sz\cm1
sort date
gen corp3=corp1 in -5/-1
gen market3=market1 in -5/-1
gen date1=date in -5/-1
sort date1
drop if corp3==.
drop date corp1 market1
save cm4,replace
clear
use cm2
sort date
gen corp3=corp2 in 1/5
gen market3= market2 in 1/5
gen date1=date in 1/5
sort date1
drop if corp3==.
drop date corp2 market2 corp market
save cm5,replace
append using cm4,keep(date1 corp3 market3)
append using cm3,keep(date1 corp3 market3)
sort date1
save cm7,replace
reg corp3 market3 in 1/100
predict stddp in 101/110
gen res= corp3-stddp in 101/110
mean res

写的真复杂，擦

maximus11111 发表于 2011-4-28 16:20 比如我要去20050510这一天前的5天和后3天的数据怎么取？

g d=date(substr(var1,1,10),"YMD")
keep if d>16560&d<16570

4# sungmoo
哈哈，早知道就好了