| Library: | panel |
| See also: | panstats pantimes |
| Quantlet: | panlag | |
| Description: | computes the lag of the variable |
| Usage: | y = panlag(z,a{, T}) | |
| Input: | ||
| z | NT x {2+}k matrix of panel data | |
| a | scalar, parameter used to compute y(i,t)=(1-a)*z(i,t)+a*z(i,t-1). if a = 0, no lag but truncation of the first observation if a = 1, lag with truncation of the last observation | |
| T | optional scalar, number of time periods for a BALANCED data set. Used for a data set with the same T for all individuals = BALANCED PANEL | |
| Output: | ||
| y | output matrix with lagged (truncated) observations. For unbalanced data: The first two columns are the adapted indices for the individuals and time periods. | |
library("panel")
z=read("dynpanel.dat") ; read data set
it=z[,1:2] ; matrix of indices
z0=panlag(it~z[,4],0) ; truncate first observation
z1=panlag(it~z[,4],1) ; lag of second variable
itlag=z0[,1:2] ; new indices
dz2=z0[,3]-z1[,3] ; differences of the second var.
itlag~dz2
Remark: to save space, only the first 16 elements of itlag and dz2 are shown here. Contents of _tmp [ 1,] 1 2 1.2226 [ 2,] 1 3 0.22922 [ 3,] 1 4 -0.29302 [ 4,] 1 5 -1.1603 [ 5,] 1 6 -0.53256 [ 6,] 1 7 0.34027 [ 7,] 1 8 -1.296 [ 8,] 1 9 0.45796 [ 9,] 1 10 1.868 [ 10,] 1 11 -1.763 [ 11,] 1 12 0.66435 [ 12,] 1 13 -1.1309 [ 13,] 1 14 -0.87567 [ 14,] 1 15 -0.39465 [ 15,] 2 2 -2.634 [ 16,] 2 3 0.78664