[R] circular regression question

[Antonio Silva]

Reading the help page with attention I saw that this function fits a
regression model for "circular dependent and linear independent".

So my question now is, is there a way to fit a model been with the circular
as the dependent variable?

Thanks

Antonio



2015-10-24 20:19 GMT-02:00 Antonio Silva <aolinto.lst em gmail.com>:

> Dear R users
>
> I'm trying to reproduce the results from Lowry et al. 2007 Lunar landings
> - Relationship between lunar phase and catch rates for an Australian
> gamefish-tournament fisheryFisheries Research 88: 15–23
>
> Basically we have two columns: Lunar  days and CPUE (catch per unit
> effort). The aim is to test whether CPUE varies with the lunar cycle
>
> Here is what I did:
>
> library(circular)
> # Black marlin CPUE
>
> U<-c(0.02,0.024,0.017,0.02,0.018,0.034,0.042,0.026,0.017,0.019,0.006,0.008,0.011,0.014,0.007,0.018,0.008,0.004,0.008,0.013,0.011,0.008,0.006,0.004,0.008,0.005,0.016,0.011,0.022,0.048)
> # Lunar day
> LD <- seq(1:30)
> # Lunar Day in radians
> LDrad <- (360*LD)/29.58
> # Plots
> plot(U~LD)
> plot(U~LDrad)
> # Transform
> LDcir <- circular(U,LDrad,type=c("angles"),units=c("radians"))
> # circular model
> circ.lm<-lm.circular(y=U,x=LDcir,init=1,type="c-l",verbose=TRUE)
>
> but it runs with
> circ.lm<-lm.circular(x=U,y=LDcir,init=1,type="c-l",verbose=TRUE)
>
> Nevertheless U must be the dependent variable, not the independent one.
>
> I also tried
>
> Uy <- cbind(U,rep(1,length(U)))
> circ.lm<-lm.circular(y=Uy,x=LDcir,init=2,type="c-l",verbose=TRUE)
>
> Well I really appreciate any help, thanks in advance,
>
> Antonio Olinto
>



-- 
Antônio Olinto Ávila da Silva
Biólogo / Oceanógrafo
Instituto de Pesca (Fisheries Institute)
São Paulo, Brasil

	[[alternative HTML version deleted]]