Transition Matrix

[Joseph Ejima]

Hello,
I am trying to generate a transition matrix of the change in the employment status of individuals across different rounds using the form below:
Round 1 Round 2
agricultural employment wage employment non farm employment Total
agricultural employment
wage employment
non farm employment
Total
I have tried the tab command for this but I keep getting an error message that "too many variables specified"
Any help on how to go about this will be highly appreciated.
Many thanks in advance.

Example generated by -dataex-. For more info, type help dataex
clear
input double indiv_id byte round float(any_work_r ag_work_r wage_work_r nfe_work_r)
1000101 1 1 0 0 1
1000101 2 1 0 0 1
1000101 3 1 0 0 1
1000101 4 1 0 0 1
1000102 1 1 0 1 0
1000102 2 1 0 1 0
1000102 3 1 0 0 1
1000102 4 0 0 0 0
1000104 1 0 0 0 0
1000104 2 0 0 0 0
1000104 3 1 0 0 1
1000104 4 0 0 0 0
1000901 1 1 1 1 0
1000901 2 1 0 1 0
1000901 3 1 0 1 0
1000901 4 0 0 0 0
1000902 1 1 1 1 0
1000902 2 1 1 1 0
1000902 3 1 1 0 0
1000902 4 1 0 0 1
1000906 1 1 1 0 0
1000906 2 0 0 0 0
1000906 3 0 0 0 0
1000906 4 0 0 0 0
1003402 1 0 0 0 0
1003402 2 1 0 0 1
1003402 3 1 1 0 0
1003402 4 1 0 0 1
1003405 1 1 0 1 0
1003405 2 1 0 1 0
1003405 3 1 0 1 0
1003405 4 1 0 1 0
1003407 1 0 0 0 0
1003407 2 0 0 0 0
1003407 3 0 0 0 0
1003407 4 0 0 0 0
1003408 1 0 0 0 0
1003408 2 0 0 0 0
1003408 3 1 0 0 1
1003408 4 0 0 0 0
1003409 1 0 0 0 0
1003409 2 0 0 0 0
1003409 3 0 0 0 0
1003409 4 0 0 0 0
1003410 1 0 0 0 0
1003410 2 0 0 0 0
1003410 3 0 0 0 0
1003410 4 0 0 0 0
1003502 1 1 0 0 1
1003502 2 1 0 0 1
1003502 3 0 0 0 0
1003502 4 0 0 0 0
1003503 1 0 0 0 0
1003503 2 0 0 0 0
1003503 3 1 0 1 0
1003503 4 0 0 0 0
1003506 1 0 0 0 0
1003506 2 1 0 1 0
1003506 3 0 0 0 0
1003506 4 0 0 0 0
1003801 1 1 0 0 1
1003801 2 1 0 0 1
1003801 3 1 0 0 1
1003801 4 1 0 0 1
1003802 1 1 0 0 1
1003802 2 1 1 0 0
1003802 3 1 0 0 1
1003802 4 1 1 0 0
1003804 1 0 0 0 0
1003804 2 0 0 0 0
1003804 3 0 0 0 0
1003804 4 0 0 0 0
1003805 1 0 0 0 0
1003805 2 0 0 0 0
1003805 3 0 0 0 0
1003805 4 0 0 0 0
1005402 1 1 0 0 1
1005402 2 1 0 1 0
1005402 3 1 0 1 0
1005402 4 1 0 1 0
1006001 1 1 0 0 1
1006001 2 1 0 0 1
1006001 3 1 1 0 0
1006001 4 1 0 0 1
1006002 1 1 0 0 1
1006002 2 1 0 0 1
1006002 3 1 0 0 1
1006002 4 1 0 0 1
1006702 1 1 1 0 0
1006702 2 0 0 0 0
1006702 3 1 1 0 0
1006702 4 1 0 0 1
1006913 1 1 1 0 0
1006913 2 1 1 0 0
1006913 3 0 0 0 0
1006913 4 1 1 0 0
1007101 1 1 1 0 0
1007101 2 1 1 0 0
1007101 3 1 0 0 1
1007101 4 1 0 0 1
end

any_work_r : employed
ag_work_r : agricultural employment
wage_work_r: wage employment
nfe_work_r: non farm employment
1= yes, 0=no

[Maarten Buis]

How do you want to handle people who work in multiple industries at the same time? For example, person 1000901 round 1.

[Joseph Ejima]

How do you want to handle people who work in multiple industries at the same time? For example, person 1000901 round 1.

That individual is actually an outlier and I will have to drop the observation. Probably there was a problem with the coding of the data as I am only interested in their main employment