[R] 3D surface plot

[David Winsemius]

On Aug 16, 2011, at 11:07 PM, Eric Heupel wrote:

> Thanks for the prompt reply. Will dig into rgl ASAP.
>
> My sample data set is:
>
>
> B,C,D,A
> 1,1,1,1
> 1,1,2,0.5
> 2,1,1,2
> 1,2,1,2
> 1,1,3,0.333333333
> 3,1,1,3
> 1,3,1,3
> 2,1,2,1
> 1,2,2,1
> 2,2,1,4
> 2,1,3,0.666666667
> 3,1,2,1.5
> 1,2,3,0.666666667
> 1,3,2,1.5
> 3,2,1,6
> 2,3,1,6
> 2,2,2,2
> 3,1,3,1
> 1,3,3,1
> 3,3,1,9
> 2,2,3,1.333333333
> 3,2,2,3
> 2,3,2,3
> 3,2,3,2
> 2,3,3,2
> 3,3,2,4.5
> 3,3,3,3

Here's a wireframe approach... which requires that we define a  
variable that holds the ratio C/D, so this is a sort of projection.  
Any case on the C/D value of 1 ( a plane in either 3 or 4 space I  
believe) will be on the CoverD line = 1 on that plot. The equi-ratio  
planes will fan out from the A=0, B=0 axis .... I think.

 > dput(dat)
structure(list(B = c(1L, 1L, 2L, 1L, 1L, 3L, 1L, 2L, 1L, 2L,
2L, 3L, 1L, 1L, 3L, 2L, 2L, 3L, 1L, 3L, 2L, 3L, 2L, 3L, 2L, 3L,
3L), C = c(1L, 1L, 1L, 2L, 1L, 1L, 3L, 1L, 2L, 2L, 1L, 1L, 2L,
3L, 2L, 3L, 2L, 1L, 3L, 3L, 2L, 2L, 3L, 2L, 3L, 3L, 3L), D = c(1L,
2L, 1L, 1L, 3L, 1L, 1L, 2L, 2L, 1L, 3L, 2L, 3L, 2L, 1L, 1L, 2L,
3L, 3L, 1L, 3L, 2L, 2L, 3L, 3L, 2L, 3L), A = c(1, 0.5, 2, 2,
0.333333333, 3, 3, 1, 1, 4, 0.666666667, 1.5, 0.666666667, 1.5,
6, 6, 2, 1, 1, 9, 1.333333333, 3, 3, 2, 2, 4.5, 3), CoverD = c(1,
0.5, 1, 2, 0.333333333333333, 1, 3, 0.5, 1, 2, 0.333333333333333,
0.5, 0.666666666666667, 1.5, 2, 3, 1, 0.333333333333333, 1, 3,
0.666666666666667, 1, 1.5, 0.666666666666667, 1, 1.5, 1)), .Names =  
c("B",
"C", "D", "A", "CoverD"), row.names = c(NA, -27L), class = "data.frame")
 >

 > require(rms)
 > dat$CoverD <- with(dat, C/D)

 > mod <- ols(A~ rcs(B,3)*rcs(CoverD), data=dat)

 > ddd <- datadist(dat)
 > options(datadist="ddd")
 > ddd <- datadist(dat)
 > bplot(Predict(mod, B, CoverD), lfun=wireframe)



>
>
>> On Aug 16, 2011, at 10:48 PM, David Winsemius wrote:
>
>>
>> On Aug 16, 2011, at 9:50 PM, Eric Heupel wrote:
>>
>>> I have what is probably a noob question, but....
>>>
>>> I am trying to create a 3d plot to illustrate the range of values  
>>> for the following simple function:
>>>
>>> A = B*(C/D)
>>>
>>> B, C, and D are independent variables whose range are equal (e.g.  
>>> 1 to 3 inclusive)
>>>
>>> I figure it's not possible to map the surface of A on the 3d space  
>>> defined by B, C and D but I would like to create a surface defined  
>>> by the lower and upper limits of the variables - that is to say a  
>>> rectangle with corners at (1,1,1), (2,3,2), (3,3,3) and (3,2,2)  
>>> with a color map displayed on it corresponding to the values of A  
>>> and a color key to the side of that.
>>>
>>> I have been able to wrap my head part way around persp and  
>>> wireframe and can create a surface for A~B*(C/D) in either, but  
>>> have not managed to create either the
>>
>> The rgl package will allow plotting in pseudo 3D space ... once you  
>> unwrap your head sufficiently to provide a suitable sample dataset.  
>> Persp and wireframe are designed for a function of two variables:   
>> A ~ B*C. You could also plot slices at various levels of D with  
>> those or perhaps more clearly with levelplot
>>
>> -- 

David Winsemius, MD
West Hartford, CT