Keywords - Function groups - @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

salsa - select0item - setmaskl - setyaxis - sijci - size - sort - SPKRprmat - SPPPinitrandom - sqrt - stockest - suitind - symweigh - systemcall
salsa Routines from time series program TRAMO
salsa call a TRAMO-DLL

salsamain loads libraries
savemac savemac saves the given quantlet into a file
scatter Computes a scatterplotmatrix for the selected variables. If two variables are chosen a simple plot is created.
scatterp Computes a scatterplotmatrix for the selected variables after performing a regression analysis.. If two variables are chosen a simple plot is created.
seats
second returns the second as a string
selec selects rows from the matrix mat
select select calculates semiparametric estimators of the intercept and slope coefficients in the "outcome" or "level" equation of a self-selection model. select is therefore the second stage of the two-stage procedure used to estimate these models. select does not estimate the coefficients of the "selection" or "decision" equation but requires that the estimated first-step "index" be given as an input. The procedure to estimate the slope coefficients is desribed in Powell (1987). The procedure to estimate the intercept coefficient is desribed in Andrews and Schafgans (1994). select combines the powell and andrews macros of the metrics library.
select0item select0item is an item selector for glm. It allows exactly at most one item to be selected.
select1item select1item is an item selector for glm. It allows exactly one item to be selected.
selectitem opens a self defined menu box to ask for a choice. The choice may have to be confirmed or not (mode="single").

selectitemlist opens a self defined menu box to ask for a choice like selectitem. For more than 15 variables a "forward" and a "backward" option ist implemented.
seq Estimates a simultaneous equations model by 3-stage least squares
setenv setenv sets one of the environment variables (xpl4home, xpl4data, xpl4help, xpl4prog, xpl4backup, format, stringformat, browser, outheadline, outlayerline, outlineno)
setfractions changes proportions of parts of display

setgopt setgopt controls the layout of a display. First create and show the display. Then call setgopt to change its headline, axes labels and limits, etc.. There are two versions of setgopt: one, where you explicitly specify the desired display options and one, where you get the display options from another display via getgopt.

setheadline setheadline changes the head line of plot.

setmask Front-end for the setting of mask vectors that allows easy definitions for points, lines (polygons), surfaces and text.
setmaskl Use setmaskl to connect datapoints by lines. setmaskl has to preceed show. That is, first you call setmaskl to specify which points to connect and the layout of the line(s). Then you use the command show to plot the datapoints and the line(s) that connect them..

setmaskp Use setmaskp to control the color, the graphical representation (symbol) and the size of each point of a datamatrix you want to plot. setmaskp has to preceed show. That is, first you call setmaskp to specify colour, graphical representation and size of the data you want to plot. Then you call show to actually plot the data.

setmaskt use setmaskt to assign labels to the points of a dataset. setmaskt has to preceed show.

That is, first you call setmaskt to assign labels to the datapoints. Then you call show

to plot the labeled data.

setmode setmode sets the mouse mode of a plot.


0=ZOOM and INDEX


1=BRUSH


2=defines your own actions using readevent

setsize setsize sets the size (in pixels) for all the plots created later

settime settime generates from a set of vectors a vector of time points.

The parameter yr (year) is necessary, the parameters mo (month), dy (day

of month), hr (hour), mn (minute), sc (second) are optional.

The granulation of the measurement can be changed with the command

settimegran.

Each value in the produced set of vectors describes the distance to a fixed point in time

(usually 1.1.1900 0:00:00.0). The zero time can be changed with the

command settimezero.

The following example teaches you to be careful not to span too much time with a too small

granulation (some kind of overflow/underflow problem).

settimegran settimegran changes the granulation. Possible values for newgran are 1...7 (Seconds, minutes, hour, days, weeks, month, years)
settimezero settimezero changes the fixed time point.

setxaxis setxaxis changes the layout of the horizontal axis.

setxgobicolormap
setyaxis setyaxis changes the layout of the vertical axis.

sfcoeff estimates standard errors of parameter estimates
sfvonbss standard errors of parameters for Subset VAR
sfvonmw standard errors for mean in VAR models
shiftr Shifts the rows of a matrix
show show is used to put graphical objects (such as a datamatrix) into a display. Before you call show you have to create a display using the command createdisplay.

showd creates global defined dipsplays which cannot be overwritten according the created parameters r c d to show the computed objects
sigma
sigma1 auxiliary quantlet for full VAR model analysis
sign Computes the sign function (0, -1, 1) for zeros, negative or positive values.
sijci auxiliary quantlet for cointegration
simdep Computes the simplicial depth estimate of location
simex SIMEX (SIMulation EXtrapolation) is a simulation-based method of estimating and reducing bias due to measurement error. simex is applicable to general estimation methods, for example, least-squares, maximum likelihood, quasi-likelihood, etc.
simpsonint simpsonint computes the integral of a given realvalued function about a d-dimensional cube [a_1,b_1] x ... x [a_d,b_d]. The method of simpson is used.
simvar computes a multidimensional autoregressive time series.
sin Returns the sine in radian of the elements of an array.
sinh Returns the hyperbolic sinus of the elements of an array.
sir Calculates the effective dimension-reduction (edr) directions by Sliced Inverse Regression (Li, 1991)
sir1 This macro is the taylored Sliced Inverse Regression method for the "sssm" macro. Here, the macro "sir1" builds a special slice which contains the cases such that y=0 (the value "0" for y symbolically indicates a missing value). Then, nbslices (by default, nbslices=5) other slices are made by splitting the range of the non missing y's into slices of nearly equal weight (i.e. such that they contain nearly the same number of cases). The classical algorithm of S.I.R. is then used with this special slicing.
sir2 Calculates the effective dimension-reduction (edr) directions by Sliced Inverse Regression II (Li, 1991)
size size gives the number of elements, containing in a list.
sker sker computes a direct kernel estimate without binning from scatter plot data.

skewness Computes the skewness for a given vector.
sknn sknn computes the k-nearest neighbour smooth regression from scatter plot data. As inputs you have to specify the explanatory variable x, the dependent variable y and the smoothing parameter k.
smoothermain loads the kernels needed by the smoother lib functions
smoothertest smoothertest tests all the aforementioned macros of the smoother.lib
smsr
softadap
softauto Softthresholds the mother wavelet coefficients b1 and b2 automatically by sqrt(2 sigma n). To compute the threshold value only b1 and x is used.
softthres Softthresholds the mother wavelet coefficients b1 and b2 interactively. The user is a threshold offered by sqrt(2 sigma n). To compute the threshold value only b1 and x is used.
sort sort sorts the rows of a matrix. If column c1 is specified the matrix will be sorted with respect to column c1. That is, the rows of the matrix will be arranged in order that elements of column c1 are in ascending (descending) order.
spatialmain Loads the dll needed for the quantlets in spatial.
spatialSPKRtest
spatialSPPPtest
spec estimates and plots the spectral density of a time series
spfill spfill fills places of sparsity with interpolated observations to avoid the need of oversmoothing.
SPKRcorrelogram Computes spatial correlograms of spatial data or residuals. Initially, it divides the range of the data into nint bins, and computes the covariance for pairs with separation in each bin, then divides by the variance. It returns results only for bins with 6 or more pairs.
SPKRexpcov Spatial covariance function for use with SPKRsurfgls.
SPKRgaucov Spatial covariance function for use with SPKRsurfgls.
SPKRmultcontours Draws multiple contour lines of a spatial object of type "trmat", "prmat", or "semat".
SPKRprmat Evaluates a Kriging surface over a grid.
SPKRsemat Evaluates a Kriging standard error of prediction surface over a grid.
SPKRsphercov Spatial covariance function for use with SPKRsurfgls.
SPKRsurfgls Fits a trend surface by generalized least squares.
SPKRsurfls Fits a trend surface, i.e., a polynomial regression surface, by least squares.
SPKRtrmat Evaluates a trend surface over a grid.
SPKRvariogram Computes spatial (semi-)variograms of spatial data or residuals. Initially, it divides the range of the data into nint bins, and computes the average squared difference for pairs with separation in each bin. It returns results only for bins with 6 or more pairs.
spline spline fits a cubic spline to input data.
SPPPgetregion Retrieves the rectangular spatial domain that previously has been set by SPPPinit or SPPPsetregion.
SPPPinit Creates a point process object and calls SPPPsetregion to set the rectangular spatial domain.
SPPPinitrandom Resets the random number generator for point processes.
SPPPkaver Computes average of simulations of K-fns.
SPPPkenvl Computes envelope (upper and lower limits) and average of simulations of K-fns.
SPPPkfn Computes K-fn of a point pattern. Actually, it computes L = sqrt(K / pi). Note that SPPPinit or SPPPsetregion must have been called before.
SPPPpsim Simulates a Binomial (Poisson) spatial point process. Note that SPPPinit or SPPPsetregion must have been called before to set the domain. To be able to reproduce results, reset the random number generator for point processes by calling SPPPinitrandom first.
SPPPsetregion Sets the rectangular spatial domain for spatial point pattern analysis.
SPPPssi Simulates a SSI (sequential spatial inhibition) point process. Note that SPPPinit or SPPPsetregion must have been called before to set the domain. To be able to reproduce results, reset the random number generator for point processes by calling SPPPinitrandom first. Note that this quantlet will never return a result if r is too large and it is impossible to place n points.
SPPPstrauss Simulates a Strauss spatial point process. It uses a spatial birth-and-death process for (4 n) steps (or for (40 n) steps when starting from a binomial pattern on the first call from another function). Note that SPPPinit or SPPPsetregion must have been called before to set the domain. To be able to reproduce results, reset the random number generator for point processes by calling SPPPinitrandom first.
sptest Additive component analysis in additive separable models using wavelet estimation. An additive component can be tested against a given polynomial form with degree p, e.g. when p is set to zero we test for significant influence of that component. The procedure is presented in Haerdle, Sperlich, Spokoiny (2000) but implemented without the "t_{j,alpha}" correction.
spur computes the trace of the matrix
sqrt sqrt computes the square root of the elements of an array.

ssr Returns the sum of squared residuals in a regression tree.
sssm computes the estimates of the slope vectors in the outcome equation and in the selection equation for a semiparametric sample selection model (SSSM). This is a two-stages estimation method: the first one is a taylored Sliced Inverse Regression (S.I.R.) analysis, and, in the second one, two Canonical analyses are conducted in order to convert the estimated EDR directions into estimates of the slopes. The identifiability conditions have to be verified: there exists a component of the explanatory variable X which affects the selection and does not affect the outcome, and there also exists another component of X which affects the outcome and does not affect the selection. If these identifiability conditions are not verified, the program stops. Moreover, the number of explanatory variables for the selection (resp. outcome) equation must be greater or equal than 2.
stack Joins two arrays along the third dimension.
standard standardizes the selected variables in ISTA. The transformed variables can replace the original ones or can be appended on the end of data.x. the type is automatically set continuous.
station test for structural change (for VAR models)
statsmain sets defaults for library stats.
statstest executes some tests for the macros defined in stats.lib. Is invoked by vertestl().
stein Stein computes the optimal threshold for a vector of data plus noise so that the mean squared error is minimized. Stein uses Stein's unbiased risk estimator for the risk. The quantlet sure uses stein to threshold the father and mother wavelet coefficients.
steps4plot produces a matrix of points for plotting a left continuous step function.
stockest stockest is estimating from a given dataset of a random process parameters for the following models: assuming a Wiener Process (model 1), assuming a compounded Poisson Jump Process mixed with a Wiener Process (model 2)
stockestsim using the given data stockestsim is first estimating the parameters for the following models: a Wiener Process (model 1) and a Wiener Process with jumps where the jumps are following a compounded Poisson Jump Process (model 2) and then comparing model 1 and model 2 with the real dataset by simulation
stocksim stocksim is simulating random processes for a stock price by three different ways: using a Wiener Process, using a compounded Poisson Jump Process with a log normal distribution of jump height and using a mixture of both
stointp Auxiliary routine for rICfil: calculates for dimension p>(=)2 diag(E[ YY' u min(b/|aIhY|,u) ]) and diag(E[ YY' min(b/|aIhY|,u)^2 ]) for u square root of a Chi^2_p-variable, and Y~ufo(S_2) indep of u by using a polar representation of Lambda:= I^{1/2} Y u, u = | I^{-1/2} Lambda |, Y=I^{-1/2} Lambda /u

The integrals are evaluated stepwise, first conditioning on Y and calculated "analytically" using Ewinn, Ew2inn and then the outer integration is done by MC-Integration along the directions Y, parametrized by a sin-cos-representation.

stointpm Auxiliary routine for rICfil: calculates for dimension p>(=)2 (E[ YY' u min(b/|aIhY|,u) ]) and (E[ YY' min(b/|aIhY|,u)^2 ]) for u square root of a Chi^2_p-variable, and Y~ufo(S_2) indep of u by using a polar representation of Lambda:= I^{1/2} Y u, u = | I^{-1/2} Lambda |, Y=I^{-1/2} Lambda /u.

The integrals are evaluated stepwise, first conditioning on Y and calculated "analytically" using Ewinn, Ew2inn and then the outer integration is done by MC-Integration along the directions Y, parametrized by a sin-cos-representation.

string string converts a set of vectors through a format string in a string.
strlen strlen returns the length of a string, i.e. the number of characters in the string.
strtok strtok breaks the string in tokens. The second argument is a set of delimiters, i.e. characters that separate token. The delimiters themselfs are not returned.
strucbru auxiliary quantlet for multi
substr substr extracts a substring from a string.
suitind Supporting Quantlet for cartsplit
sum sum computes the sum of the elements of an array regarding a given dimension.
summarize provides a short summary table (min, max, mean, median standard error) for all columns of a data matrix. An additional vector of name strings can be given to identify columns by names.
supsmo calculates the super smoother
sur Estimates a seemingly unrelated regression system by feasible generalized least squares
sure Sure denoises wavelet coefficients so that the mean squared error is minimized. MSE is estimated by Stein's unbiased risk estimator based on the variance of the coefficients. Sure computes the optimal threshold for the father wavelets and each level of mother wavelets. The input arrays can be obtained by the function 'fwt'.
sure2d Sure denoises wavelet coefficients. If the stein procedure is chosen, the mean squared error is minimized. MSE is estimated by Stein's unbiased risk estimator based on the variance of the coefficients. Sure computes then the optimal threshold for the father wavelets and each level of mother wavelets. If soft or hard thresholding is chosen, only the mother wavelet coefficients will be denoised.
svd computes the singular value decomposition of a n x p matrix x (n >= p). The singular value decomposition finds matrices u, l, v such that x = u*l*v' u, v are orthogonal matrices l is a diagonal matrix
switch switch allows the selection of one or more alternatives of many. Each alternative is introduced by a case statement. Similar to if-endif it controls, whether the following block is processed or not. The keyword break serves as end marker of case and leaves the switch block at the position of endsw. When break is omitted, the next consecutive case is processed. If the program's counter comes to default, the following block is executed in any case.

symroot Calculates the symmetric root of a symmetric positive semidefinite matrix,(s.p.s.d.) ie. symroot(x)=symroot(x)' and symroot(x)*symroot(x) = x. uesful for simulation of multivariate normal variates with a given Covariance Structure
symweigh symweigh computes the symmetrical weights
systemcall systemcall passes a UNIX command to the underlying UNIX shell.

Keywords - Function groups - @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

(C) MD*TECH Method and Data Technologies, 21.9.2000