function that generates time models and saves them in workspace.
generateTimeModels(genClasses = NULL, noGenClasses = NULL, nSam = 1, test = FALSE,
debug.level = 0)list of particular classes for which time models should be generated
list of particular classes for which time models should not be generated
number of attempts to be tried for each combination of predictions and observations, defaults to 1, higher number should be used for better accuracy. nSam/2 is used for copulas, to reduce computation time.
logical; if true, the time models are generated based on fewer iterations, for speed
if debug.level >= 1, the function will store
tables with the prediction times for each model in the workspace.
The function generates a timeModels object, which can be used to estimate
prediction times for different requests to the interpolate function
in the intamap-package, via predictTime.
This function calculates a time model for different interpolation types
in the intamap-package and returns a list object
with the estimated models. It's users responsibility to store the model in the
workspace. The normal procedure would be to run the function without arguments. However,
it is both possible to define a list for which classes the user want to
generate models, or a list of classes that are not of interest.
The time model is based on creation of a set of synthetical data sets
of different size, both regarding number of observations and prediction locations.
The function will estimate parameters and make predictions with the different combinations,
and for each method, fit a local polynomial regression model (loess)
This model can then be used by predictTime to estimate the
prediction time for an interpolation request with a certain number
of observations and prediction locations.
Pebesma, E., Cornford, D., Dubois, G., Heuvelink, G.B.M., Hristopulos, D., Pilz, J., Stohlker, U., Morin, G., Skoien, J.O. INTAMAP: The design and implementation f an interoperable automated interpolation Web Service. Computers and Geosciences 37 (3), 2011.
# NOT RUN {
timeModels=generateTimeModels()
q("yes")
## restart R in the same directory
# }
Run the code above in your browser using DataLab