I have to perform a survival analysis using data with repated observations: for each country I collected some variables once a year for a period of 30 years (1990-2020).
If in that time range the event happens, then it stays to 1 for the remaining observations. Covariates are always the same and change over time.
Here is a sample:
COUNTRY | YEAR | EVENT | AGENCY | GDP | EXPENSES | EDUCATION | TRADE |
Albania | 2017 | 0 | 1 | 37.61 | 1.11 | 57.38 | 0 |
Albania | 2018 | 0 | 1 | 40.08 | 1.17 | 54.96 | 0 |
Albania | 2019 | 0 | 1 | 41.71 | 1.27 | 59.78 | 0 |
Algeria | 1990 | 0 | 0 | 179.15 | 1.46 | 10.29 | 0 |
Algeria | 1991 | 0 | 0 | 182.99 | 1.21 | 11.01 | 0 |
Algeria | 1992 | 0 | 0 | 190.15 | 2.14 | 11.15 | 11812846 |
Algeria | 1993 | 0 | 0 | 190.57 | 2.51 | 10.99 | 10236932 |
Array
This is not readable. How can I show only the years after 1990? I tried with axis scale property, but it does not work. Any suggestions?
Now I would like to run Cox hazard regression in order to weigh the coviariates that determine risk. Is it the appropriate model?
Should I use any precaution, given the repeated observations?
Thanks,
0 Response to Survival analysis: repeated observations
Post a Comment