segmented linear regression

[lillymaginta]

I have the following database: I know that the data points are not sufficient to producec an output but here is the scenario I have: 

id   date          x_1 x_2 x_3
1   1/1/2001   1       0      0
1   1/2/2001   1       0      0
1  2/2/2001     0      1      0
1  3/2/2001     1      1      0
2   1/1/2001   1       0      0
2  1/2/2001    1       0      0
2  2/2/2001     0      1      0
2  3/2/2001     1      1      0
1   1/1/2002   1       0      1
1   1/2/2002   1       1      0
1  2/2/2002     1     1      0
1  3/2/2002    1      1      0
1   7/1/2002   1       0      1
1   11/2/2002  1       1      0
1  32/2002     1     1      0
1  3/2/2002    1      1      0

I have two periods 1/2001-12/2001 and 1/2002-12-2002. I am trying to compare the slopes of each period for each of the x. So I am assuming I will build a speperate model for x1, x2, x3. Each id could have multiple entries per month but my goal is to compare the cumulative change across al ids in the first period compared to the second period. Any help is appreciated  

[Ksharp]

It is more like time series's PANEL data. Could you post it at Time Series Analysis forum? Someone could you some advice .