pairid | tumor | tumortype |
1 | no | non |
1 | no | non |
1 | no | non |
1 | no | non |
1 | yes | primary |
2 | no | non |
2 | no | non |
2 | no | non |
2 | no | non |
2 | yes | metastatic |
From this main dataset, I would like to analyze the subset of only primary tumors with their matched non tumor controls. There are 825 primary tumors. Primary tumors are coded 0, metastatic tumors are coded 1.
When I run commands such as
. graph box age if tumortype==0, over(tumor)
I wanted to see two boxes, one for tumor (in this case primary only) and one for their matched non tumors. In reality, I only see one box labeled tumor.
or
. clogit tumor age if tumortype==0, group(pairid)
yields: outcome does not vary in any group
or
. by tumortype, sort : clogit tumor age, group(pairid)
yields: -> tumortype = primary
outcome does not vary in any group
Any help would be appreciated.
Thanks
Richard
0 Response to Analyze a subset of a matched case control dataset
Post a Comment