[R] Getting predictions for skewed t-distribution

[Patrick Connolly]

I have a small dataframe xxF, a summary of which looks like this:

> summary(xxF)
       T              Dev          
 Min.   :10.44   Min.   :0.008929  
 1st Qu.:10.44   1st Qu.:0.012048  
 Median :18.61   Median :0.031250  
 Mean   :17.87   Mean   :0.028286  
 3rd Qu.:22.24   3rd Qu.:0.041667  
 Max.   :30.37   Max.   :0.050000  


I managed to make a non-linear fit after a lot of fiddling with
initial values but it looks overly complicated and biologically
unconvincing in part.  The general form of a skewed t-distribution
looks more appropriate so I tried selm from the sn package thus:


>  selmFt <- with(xxF, selm(Dev ~ T, family = "ST", method="MPLE"))
>  coef(selmFt, param.type="DP")
(Intercept.DP)              T          omega          alpha             nu 
  -0.015895099    0.002689226    0.002306132   -5.660870446    1.473210455 

I wish to get predictions for values of T between 10 and 32 but I can't
figure out how to use those coefficients.

With an linear model or glm, even without a prediction method, it's
fairly simple to get predictions from a range of values of the
independent variable/s.  For a skewed-t it's evidently less
straightforward.  Does it have to be done using CP type parameters?


Ideas gratefully accepted



In case it makes any difference....

> sessionInfo()
R version 3.2.1 (2015-06-18)
Platform: i686-pc-linux-gnu (32-bit)

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats4    grDevices utils     stats     graphics  methods   base     

other attached packages:
[1] dplyr_0.3.0.2      nlmrt_2013-9.25    RColorBrewer_1.1-2 plyr_1.8.3        
[5] stringr_1.0.0      reshape2_1.4.1     sn_1.2-2           lattice_0.20-31   

loaded via a namespace (and not attached):
 [1] Rcpp_0.11.3       assertthat_0.1    grid_3.2.1        DBI_0.3.1        
 [5] magrittr_1.0.1    stringi_0.4-1     lazyeval_0.1.10   tools_3.2.1      
 [9] numDeriv_2014.2-1 parallel_3.2.1    mnormt_1.5-3 


-- 
~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.   
   ___    Patrick Connolly   
 {~._.~}                   Great minds discuss ideas    
 _( Y )_  	         Average minds discuss events 
(:_~*~_:)                  Small minds discuss people  
 (_)-(_)  	                      ..... Eleanor Roosevelt
	  
~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.