I have time series data for 10 years ( hourly, daily, monthly ) for different variables , i have large numbers of missing data which are not random missing so how can i apply GBM or linear regression in R to replace missing data
i have got this code from this site: http://s3l.stanford.edu/blog/?p=73
what is y ( is it years???) , and that are independent variables
still it is not clear to me, i need your help
gbm1
Dear Safia,
You may use gbmImpute() function from package 'imputation'. It is very straightforward, but not on CRAN now (can be installed from the archived source file). AFAIK, its first parameter is a data.frame containing NAs, and returns the same data.frame imputed with predicted values.
HTH,
Ákos
many thanks
but still i have the same problem
i mean if i used regression, it became more easier but still i am confusing
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series