monreg provides a strictly monotone estimator of the regression function based on the nonparametric regression model.
monreg(x, y, a = min(x), b = max(x), N = length(x), t = length(x),
hd, Kd = "epanech", hr, Kr = "epanech", degree = 1,
inverse = 0, monotonie = "isoton")
vector containing the x-values (design points) of a sample
vector containing the y-values (response) of a sample
lower bound of the support of the design points density function, or smallest fixed design point
upper bound of the support of the design points density function, or largest fixed design point
number or vector of evaluation points of the unconstrained nonparametric regression estimator (e.g. Nadaraya-Watson estimator)
number or vector of points where the monotone estimation is computed
bandwith of kernel \(K_d\) of the density estimation step
Kernel for the density estimation step (monotonization step). 'epanech' for Epanechnikov, 'rectangle' for rectangle, 'biweight' for biweight, 'triweight' for triweight, 'triangle' for triangle, 'cosine' for cosine kernel
bandwith of kernel \(K_r\) of the regression estimation step.
Kernel for the regression estimation step (unconstrained estimation). 'epanech' for Epanechnikov, 'rectangle' for rectangle, 'biweight' for biweight, 'triweight' for triweight, 'triangle' for triangle, 'cosine' for cosine kernel.
Determines the method for the unconstrained estimation.
'0' for the classical Nadaraya-Watson estimate, '1' for the local linear estimate.
As well degree can be the vector of the unconditional estimator provided by the
user for the design points given in the vector N
For '0' the original regression function is estimated, for '1' the inverse of the regression function is estimated.
Determines the type of monotonicity. 'isoton<U+00B4> if the regression function is assumed to be isotone, 'antinton' if the regression function is assumed to be antitone.
monreg
returns a list of values
the input values x, standardized on the interval \([0,1]\)
input variable y
the points, for which the unconstrained function is estimated
the points, for which the monotone function values will be estimated
length of the vector x
length of the vector z
length of the vector t
bandwidth used with the Kernel \(K_d\)
bandwidth used with the Kernel \(K_r\)
kernel used for the monotonization step
kernel used for the initial unconstrained regression estimate
method, which was used for the unconstrained regression estimate
length of the vector degree. If ldeg.vektor is not equal to 1 the user provided the vector of the unconditional estimator for the design points given in the vector N
indicates, if the origin regression function or its inverse has been estimated
the monotone estimate at the design points \(t\)
Nonparametric regression models are of the form \(Y_i = m(X_i) + \sigma(X_i) \cdot \varepsilon_i\),
where \(m\) is the regression funtion and \(\sigma\) the variance function.
monreg
performs a monotone estimate of the unknown regression function
\(m\). monreg
first estimates \(m\) by an unconstrained nonparametric
method, the classical Nadaraya-Watson estimate or the local- linear estimate
(unless the user decides to pass his or her own estimate).
In a second step the inverse of the (monotone) regression function is calculated,
by monotonizing this unconstrained estimate. With the above notation and
\(\hat m\) for the unconstrained estimate, the second step writes as follows,
$$\hat m_I^{-1} = \frac{1}{Nh_d} \sum\limits_{i=1}^N \int\limits_{-\infty}^t K_d \Bigl( \frac{\hat m (\frac{i}{N} ) - u}{h_d} \Bigr) \; du.$$
Finally, the monotone estimate achieved by inversion of \(\hat m_I^{-1}\).
monvardiff
and monvarresid
for monotone variance function estimation.
# NOT RUN {
x <- rnorm(100)
y <- x + rnorm(100)
mon1 <- monreg(x, y, hd = .5, hr = .5)
plot(mon1$t, mon1$estimation)
# }
Run the code above in your browser using DataLab