I am undertaking a dose-response meta-analysis using fractional polynomial regression model.
Below is the command I used
mfp: regress logrr dose [aweight=wt], vce(robust)
*dose refers the exposure variable and ranges between 3.6 to 96.2 (both integer and decimals).
*Afer the xblc command, I have got error messages
capture drop pa rr lb ub
quietly levelsof dose, local(levels)
quietly xblc Idose__1, covname(dose) at(`r(levels)') reference(3) eform line
The first error message is:
value 4 of dose not observed
r(198);
While the second one is below but it comes after the dose level of 9 (after I removed all from 3.6 to 9 as recommended by the above stata error message).
estimates post: matrix has missing values
r(504);
Please, assist me to solve these errors.
I thank you for your great support.
Sincerely
Nigussie
Below is the command I used
mfp: regress logrr dose [aweight=wt], vce(robust)
*dose refers the exposure variable and ranges between 3.6 to 96.2 (both integer and decimals).
*Afer the xblc command, I have got error messages
capture drop pa rr lb ub
quietly levelsof dose, local(levels)
quietly xblc Idose__1, covname(dose) at(`r(levels)') reference(3) eform line
The first error message is:
value 4 of dose not observed
r(198);
While the second one is below but it comes after the dose level of 9 (after I removed all from 3.6 to 9 as recommended by the above stata error message).
estimates post: matrix has missing values
r(504);
Please, assist me to solve these errors.
I thank you for your great support.
Sincerely
Nigussie
