interpolation option

[dvanre]

In the interpolation option 10 to 11 what is exactly meant by interpolation migrated at end/start, backward/forward extrapolation

[Antti-L]

From the documentation:
"Migration means that data points are interpolated and extrapolated within each period but not across periods. This method thus migrates any data point specified for other than milestoneyr year to the corresponding milestoneyr year within the period, so that it will be effective in that period."

Isn't this quite clear? If not, can you explain what is unclear in it?
I know the English is a bit bad, sorry about it.

Option 10 includes only migration. Option 11 additionally interpolates the time-series in the standard way (irrespective of periods), and so the migration is effective only at the start and end of the time-series.  Consequently, option 11 differs from option 1 only by migrating the first data point to the milestone year of that period, and the last data point to the milestone year of that period.

[dvanre]

If I understand correctly, when there are two or more datapoints within a period they are interpolated or extrapolated for the milestoneyear of the period. How is the extrapolation done, What does migration to the endpoints or to ends mean

Thank you

denise

[Antti-L]

As I already tried to explain in my previous answer, when the option 10 is used, if there are any datapoints within a given period, these datapoints are interpolated and extrapolated within that period.

The interpolation of datapoints means that if the milestone year is between two datapoints, the value at the milestone year is derived from those two datapoints by interpolation (I assume that you know what is meant by interpolation?).

The extrapolation of datapoints means that if the milestone year is not between any datapoints, but either before the first datapoint or after the last datapoint, the value at the milestone year is derived by extrapolating either the first databoint backwards or the last datapoint forwards. Backwards extrapolation within the period means that the value of the first datapoint is copied to the preceding milestone year in that period, and forward extrapolation means that the value of the last datapoint is copied to the succeeding milestone year in that period.

As I already explained in my previous answer, the migration at the start and end of the time series (with option 11) simply means that if the milestone year of the first or last period overlapping with the time series is not between any of the datapoints, the first datapoint is extrapolated backwards to the milestone year in that period, and the last datapoint is extrapolated forwards to the milestone year of that period. If the milestone year is between any of the datapoints, the value is, of course, derived by interpolation, as usual.

If this ansewer still does not help, could you please be more specific about the unclear details.