Hi everybody,
I try to estimate values for cells with NA by interpolation across a 24-band
layerstack (ndvi time series).
Each band is the ndvi of a monotemporal scene and is sorted by the acquisition
day in the year. The temporal distances between the scenes are different.
I use the function "approxNA" from the "raster" package. However the function
use the band number for the x-axis to determine the missing values. In this
case are the distances between the layer equal and the estimated value for a
cell is just the mean from the cell values in the layers before and behind.
Because the distances between the scenes are very different (vary from 2 to 30
days), I need a more exact estimation.
It is possible to use the real distance (day in the year) for the interpolation
with "approxNA" like in the normal "approx" function?
Thanks and best regards,
Tobias
[[alternative HTML version deleted]]
_______________________________________________
R-sig-Geo mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
