07-03-2013, 05:56 AM
If you include yearly data, there is nothing left for TIMES to do by interpolation/extrapolation, because your data is already specified for the Milestone years.
Typically, the time series data used as input for TIMES represent future projections, which usually don't have any abrupt jumps, but are locally close to linear. In these cases you don't need to do any interpolation by yourself. However, if your input data includes time series that would be far from linear within some periods, you could use e.g. 5-year centered averages for converting the yearly data into a time series with data points having 5-year intervals. Optionally, you could use discount factors as weights in the calculation to get weighted 5-year centered averages.
Typically, the time series data used as input for TIMES represent future projections, which usually don't have any abrupt jumps, but are locally close to linear. In these cases you don't need to do any interpolation by yourself. However, if your input data includes time series that would be far from linear within some periods, you could use e.g. 5-year centered averages for converting the yearly data into a time series with data points having 5-year intervals. Optionally, you could use discount factors as weights in the calculation to get weighted 5-year centered averages.