spsi (version 0.1)

sps_fun: Shape Preserving Bi-variate and Uni-variate Interpolation.

Description

Both functions recognize whether uni-variate or bi-variate interpolation is requested

sps_fun calls sps_prep to prepare spline object and according to data provided creates uni-variate or bi-variate executable function. Parameters can be set by specifying them as arguments

sps_interpolate prepares spline and returns its values on given set of tabulation points

Usage

sps_fun(x, y, z=NULL, der=0, der.x=der, der.y=der, grid = FALSE, ...)
sps_interpolate(x, y, z=NULL, xt, yt=NULL, grid = TRUE, ...)

Arguments

x

Vector of x-coordinates of data.

y

Vector of y-coordinates of data. For uni-variate case length should be equal to length of x and $y[i] = f(x[i])$. For bi-variate case length may differ.

z

Bi-variate case only. Matrix of values of function on grid spanned by x and y; $z[i,j] = f(x[i],y[j])$ OR Vector of values of functions on points (x, y); $z[i] = f(x[i],y[i])$. In this case length of all three vectors should be equal and it should be possible to transform the point to a gridded form.

xt, yt

Vectors OR Matrices containing coordinates of the tabulation point; yt for bi-variate case only

der, der.x, der.y

Indicate which derivative should be returned; der.y used in bi-variate case only. If length of vector is greater than 1 function will return list of values. For bi-variate case vectors should have equal length. At the moment only 1 and 0 are supported. See examples.

grid

For bi-variate interpolation only. If TRUE function will return matrix of values of the spline on grid spanned by vectors of tabulation points. If FALSE vector of f(xt[i] , yt[i]), i = 1, 2, ..., length(xt) will be returned. If matrices were given as tabulation points grid is meaningless.

...

arguments in tag = value format. The tags must come from the names of parameters of sps_prep and/or sps_eval.

Value

sps_fun

Function of 1 or 2 variables depending on data provided. Function will accept vector(s) or matrix(es) of tabulation point and return object of the same class. If while calling sps_fun length der.x (and possibly der.y) was bigger than 1. Resulting function will be returning list of values of respective derivatives at given tabulation points.

sps_interpolate

Vector, matrix or list of vectors/matrices of value of function and/or derivatives on given set of tabulation points

Details

Following parameters can be specified:

fx, fy, fxy

Matrices with values of the derivatives: see sps_prep

maxdeg

Maximum degree of polynomial allowed: see sps_prep

smoothness

Smoothness required: see sps_prep

tol

Relative tolerance used by program

shape

Vector of shape attributes that must be preserved. Must contain only 'monotonicity' and/or 'curvature'

References

Costantini, P; Fontanella, F; 'Shape Preserving Bi-variate Interpolation' SIAM J NUMER. ANAL. Vol. 27, No.2, pp. 488-506, April 1990

See Also

sps_eval sps_prep

Examples

## Example 1

# Following example shows usage of sps_fun along with the parameter 'smoothness'.
# As you will see if smoothness = 2 then first derivative of function is differentiable
# everywhere.

x <- c( 1, 2, 3, 4, 5, 6)
y <- c(16 ,18, 21, 17, 15, 12)

evalK1 <- sps_fun(x, y)
derK1 <- sps_fun(x, y, der.x=1)

evalK2 <- sps_fun(x, y, smoothness = 2)
derK2 <- sps_fun(x, y, smoothness = 2, der.x = 1)

xs <- seq(1, 6, 0.01)
par(mfrow = c(2, 2))
plot(x, y, col = "red", xlim = c(0, 7), ylim = c(10, 22),
     main = "Spline, smoothness = 1")
grid()

lines(xs,evalK1(xs), col="cyan")
par(new = TRUE)
plot(derK1, from = 1,to = 6, col = "magenta", xlim = c(0,7), ylim = c(-6,5),
    xaxt = 'n',yaxt='n',ann = FALSE)
axis(4, -6:5)


plot(x, y, col="red", xlim=c(0,7), ylim=c(10,22),
     main = "Spline, smoothness = 2")
grid()

lines(xs,evalK2(xs), col="cyan")
par(new = TRUE)
plot(derK2, from = 1, to = 6, col = "magenta", xlim = c(0,7), ylim = c(-6,5),
     xaxt = 'n',yaxt = 'n', ann = FALSE)
axis(4, -6:5)

plot(derK1, from = 1.5, to = 2.5)
plot(derK2, from = 1.5, to = 2.5)

## EXAMPLE 2
par(mfrow = c(1,1))
X <- seq(0, 50, 5)
Y <- seq(0, 40, 5)

X_ <- seq(0, 50, 0.5)
Y_ <- seq(0, 40, 0.5)

persp3D(X_, Y_, sps_interpolate(X, Y, akima, X_, Y_,
        grid = TRUE, shape = 'monotonicity'))