CODEWITHSUNDEEP
rinterpolationcubic-spline
Jonathan
Jonathan

Reputation: 99

View interpolated values of cubic spline (R)

I want to perform cubic spline interpolation : 

x <- c(1,1.5, 2, 4,5,6,7,8,9,10,12,15,20, 25,30)
y <- c(-0.402, -0.303, -0.198, 0.211,0.4133,0.606,0.7835,0.9404,1.0733,1.189,1.383,1.594,1.812,1.893,1.922)

smooth.spline (x,y)

you can notice that in the x vector , the x-value of 3 needs to be interpolated. How can I view interpolated values ( or insert an x values and get the interpolated y value), and view the spline functions?

Upvotes: 1

Views: 713

Answers (1)

Heikki
Heikki

Reputation: 2254

You can use predict to apply the fitted spline function:

> x <- c(1,1.5, 2, 4,5,6,7,8,9,10,12,15,20, 25,30)
> y <- c(-0.402, -0.303, -0.198, 0.211,0.4133,0.606,0.7835,0.9404,1.0733,1.189,1.383,1.594,1.812,1.893,1.922)
> predict(smooth.spline (x,y), 3)
$x
[1] 3

$y
[1] 0.006938673

Some more information on the spline functions you can get using str function:

> str(smooth.spline(x,y))
List of 19
 $ x         : num [1:15] 1 1.5 2 4 5 6 7 8 9 10 ...
 $ y         : num [1:15] -0.403 -0.301 -0.199 0.212 0.413 ...
 $ w         : num [1:15] 1 1 1 1 1 1 1 1 1 1 ...
 $ yin       : num [1:15] -0.402 -0.303 -0.198 0.211 0.413 ...
 $ tol       : num 9e-06
 $ data      :List of 3
  ..$ x: num [1:15] 1 1.5 2 4 5 6 7 8 9 10 ...
  ..$ y: num [1:15] -0.402 -0.303 -0.198 0.211 0.413 ...
  ..$ w: num 1
 $ no.weights: logi TRUE
 $ lev       : num [1:15] 0.766 0.392 0.592 0.708 0.572 ...
 $ cv.crit   : num 6.3e-06
 $ pen.crit  : num 7.17e-06
 $ crit      : num 6.3e-06
 $ df        : num 10.9
 $ spar      : num 0.336
 $ ratio     : num 5.29e-06
 $ lambda    : num 5.51e-06
 $ iparms    : Named int [1:5] 1 0 13 0 NA
  ..- attr(*, "names")= chr [1:5] "icrit" "ispar" "iter" "" ...
 $ auxM      : NULL
 $ fit       :List of 5
  ..$ knot : num [1:21] 0 0 0 0 0.0172 ...
  ..$ nk   : int 17
  ..$ min  : num 1
  ..$ range: num 29
  ..$ coef : num [1:17] -0.403 -0.369 -0.301 -0.096 0.145 ...
  ..- attr(*, "class")= chr "smooth.spline.fit"
 $ call      : language smooth.spline(x = x, y = y)
 - attr(*, "class")= chr "smooth.spline"

Upvotes: 3

Related Questions