l1pmreg {ftnonpar} | R Documentation |
Applies the generalized taut string method to quantile regression.
quantpmreg(y, beta = 0.5, squeezing.factor = 0.5, verbose = F, localsqueezing = T, DYADIC = T, thr.const = 2, extrema.nr = -1, bandwidth = -1) l1pmreg(y, squeezing.factor = 0.5, verbose = F, localsqueezing = T, DYADIC = T, thr.const = 2, extrema.nr = -1, bandwidth = -1)
y |
observed values (ordered by value of independent variable) |
beta |
quantile. The default is 0.5 which corresponds to the robust taut string. |
squeezing.factor |
The amount of decrement applied to the bandwidthes |
verbose |
logical, if T progress (for each iteration) is illustrated grahically |
localsqueezing |
logical, if T (default) the bandwidth is changed locally. |
DYADIC |
logical, if T (default) the multiresolution criterion is only verified on intervals with dyadic endpoints. |
thr.const |
smoothing parameter for the multiresoultion criterion (should be approximately 2) |
extrema.nr |
if set to a positive integer an approximation with the specified number of local extreme values is calculated |
bandwidth |
if set to a positive value the specified bandwidth is used instead of the multiresolution criterion. |
A list with components
y |
The approximation of the given data |
lambda |
The final values of lambda |
nmax |
Number of local extreme values |
Arne Kovac Arne.Kovac@uni-essen.de
D"umbgen, L. and Kovac, A. (2003) Extensions of smoothing via taut strings