我有一个面板数据集。我的变量标识符是cc代表国家/地区代码，Year代表年份：
clear
input long cc float(sch Year)
2 0 1960
2 0 1961
2 0 1962
2 0 1963
2 0 1964
2 0 1965
2 0 1966
2 0 1967
2 0 1968
2 0 1969
2 0 1970
2 0 1971
2 0 1972
2 0 1973
2 0 1974
2 0 1975
2 0 1976
2 0 1977
2 .733902 1978
2 .7566 1979
2 .78 1980
2 .875 1981
2 .9225 1982
2 1.0174999 1983
2 1.0649999 1984
2 1.16 1985
2 1.2425 1986
2 1.28375 1987
2 1.36625 1988
2 1.4075 1989
2 1.49 1990
2 1.5825 1991
2 1.62875 1992
2 1.72125 1993
2 1.7675 1994
2 1.86 1995
2 1.935 1996
2 1.9725 1997
2 2.0475001 1998
2 2.085 1999
2 2.16 2000
2 2.27 2001
2 2.325 2002
2 2.435 2003
2 2.49 2004
2 2.6 2005
2 2.7575 2006
2 2.83625 2007
2 2.99375 2008
2 3.0725 2009
2 3.23 2010
2 3.15125 2011
2 3.190625 2012
2 3.1709375 2013
2 3.1807814 2014
2 3.1758595 2015
2 3.1783204 2016
2 3.17709 2017
2 3.177705 2018
4 0 1960
4 0 1961
4 0 1962
4 0 1963
4 0 1964
4 0 1965
4 0 1966
4 0 1967
4 0 1968
4 0 1969
4 0 1970
4 0 1971
4 0 1972
4 0 1973
4 0 1974
4 0 1975
4 0 1976
4 0 1977
4 4.657455 1978
4 4.8015 1979
4 4.95 1980
4 5.4 1981
4 5.625 1982
4 6.075 1983
4 6.3 1984
4 6.75 1985
4 7.02 1986
4 7.155 1987
4 7.425 1988
4 7.56 1989
4 7.83 1990
4 7.8275 1991
4 7.82625 1992
4 7.82375 1993
4 7.8225 1994
4 7.82 1995
4 8.195 1996
4 8.3825 1997
4 8.7575 1998
4 8.945 1999
4 9.32 2000
4 9.412499 2001
4 9.45875 2002
4 9.55125 2003
4 9.5975 2004
4 9.69 2005
4 9.73 2006
4 9.75 2007
4 9.79 2008
4 9.81 2009
4 9.85 2010
4 9.83 2011
4 9.84 2012
4 9.835 2013
4 9.8375 2014
4 9.83625 2015
4 9.836875 2016
4 9.836563 2017
4 9.83672 2018
end
我想插值sch变量以减少年份。变量sch对1979-2018年的观测。通过使用对1978的观察，我想对1977的值进行插值：
sch_1977 = 0.97 * sch_1978
我尝试过的代码如下：
forvalues y = 1977 1976 1975{
local i = `y' - 1958
bysort cc (Year): generate sch`y' = 0.97*sch[`i']
replace sch`y' = 0 if Year != `y'
replace sch = sch + sch`y'
}
此处i对应于变量1978放置cc的年份的行。通过使用forvalues循环，在每个迭代中，我想创建一个新变量（sch1977，sch1978，sch1979），并在相应的年份中进行插值观测，并为零所有其他观察。接下来，我想用sch总结这个新变量。但是，Stata抱怨该代码无效。
答案 0 :(得分：2)
以下对我有用：
foreach x in 1977 1976 1975 {
local i = (2018 - 1960) - (2018 - `x') + 2
bysort cc (Year): generate sch_`x' = 0.97 * sch[`i']
replace sch_`x' = 0 if Year != `x'
replace sch = sch + sch_`x'
}
结果：
bysort cc (Year): list if inrange(Year, 1970, 1980), sepby(cc)
-> cc = 2
+-------------------------------------------------------+
| cc sch Year sch_1977 sch_1976 sch_1975 |
|-------------------------------------------------------|
11. | 2 0 1970 0 0 0 |
12. | 2 0 1971 0 0 0 |
13. | 2 0 1972 0 0 0 |
14. | 2 0 1973 0 0 0 |
15. | 2 0 1974 0 0 0 |
16. | 2 .6698126 1975 0 0 .6698126 |
17. | 2 .6905284 1976 0 .6905284 0 |
18. | 2 .7118849 1977 .7118849 0 0 |
19. | 2 .733902 1978 0 0 0 |
20. | 2 .7566 1979 0 0 0 |
21. | 2 .78 1980 0 0 0 |
+-------------------------------------------------------+
-> cc = 4
+-------------------------------------------------------+
| cc sch Year sch_1977 sch_1976 sch_1975 |
|-------------------------------------------------------|
11. | 4 0 1970 0 0 0 |
12. | 4 0 1971 0 0 0 |
13. | 4 0 1972 0 0 0 |
14. | 4 0 1973 0 0 0 |
15. | 4 0 1974 0 0 0 |
16. | 4 4.250733 1975 0 0 4.250733 |
17. | 4 4.382199 1976 0 4.382199 0 |
18. | 4 4.517731 1977 4.517731 0 0 |
19. | 4 4.657455 1978 0 0 0 |
20. | 4 4.8015 1979 0 0 0 |
21. | 4 4.95 1980 0 0 0 |
+-------------------------------------------------------+