pmden: Piecewise monotone density estimation with taut strings
In ftnonpar: Features and Strings for Nonparametric Regression

Applies the taut string method to one-dimensional data.

1	pmden(x, DISCR=FALSE,verbose = FALSE, bandwidth=-1, extrema.nr = -1, accuracy = mad(x)/1000, extrema.mean = TRUE,maxkuipnr=19,asympbounds=FALSE, tolerance = 1e-08,localsq=TRUE,locsq.factor=0.9)

`x`	observed values
`DISCR`	logical, if T a discrete density is fitted
`verbose`	logical, if T progress (for each iteration) is illustrated grahically
`bandwidth`	if set to a positive value the specified bandwidth is used instead of the automatic criterion based on generalized Kuiper metrics.
`extrema.nr`	if set to a positive integer an approximation with the specified number of local extreme values is calculated
`accuracy`	Precision of the data. Handling of identical observations depends on this parameter.
`extrema.mean`	logical, if T the value at the local extrema is changed to the mean frequency of observations on that interval
`maxkuipnr`	The order of the highest generalized Kuiper metric used for the automatic choice of the bandwidth
`asympbounds`	If set to T asymptotic bounds derived from a Brwonian Bridge are used for the Kuiper criterion. Otherwise simulated bounds for various sample sizes are interpolated for the size of the data x
`tolerance`	Accuracy used for the determination of the bandwidth when extrema.nr is greater than 0.
`localsq`	If set to TRUE (default) performs, if necessary, additional local squeezing after the Kuiper metrics are small enough
`locsq.factor`	The amount of decrement applied to the bandwidthes if local squeezing is carried out.

`y`	values of the density approximation between the observations
`widthes`	bandwidth used for the taut string approximation
`nmax`	number of local extreme values
`ind`	indices of knots points of the taut string
`trans`	taut string at the observations, should look like uniform noise