Re: [R] posting for R-help on the predict function and categorical variables in R
How predict works depends on the method written for that type of object. The zeroinfl function is not in any of the standard packages, so it must be in another package, but you did not tell us which. Since it is from a package other than the main ones, it may work similarly to the regular predict functions, or it may work completely different. Have you tried reading the help pages for zeroinfl and the predict method for it? Have you worked through any vignettes for that package? On Fri, Mar 7, 2014 at 10:16 AM, Roman, Sally (MRC) sally.ro...@mrc.virginia.gov wrote: Hi - I would like to post to the R-help mailing list. Here is my post: -This is more of a general question about how the predict function treats categorical variables and how to interpret the output from predict. I have a zeroinfl model to predict the number of animals encountered: b9-zeroinfl(Count ~ as.factor (Area) + as.factor(Season)|1, dist=negbin,data = total) where Count is the number of animals and the explanatory variables are Area and Season, both are coded as factors in the model. Area has three levels and Season has 4 levels. Coding the two as factors allows for R to create dummy variables for each variable for use in the model. When I use the predict function to predict the number of animals for a larger data set, I want to make sure I'm understanding what is happening. My newdata for predict is: newdata- as.data.frame(Season, Area) Both variables are coded as factors and the dataframe is in a long format. There are records for each combination of Season and Area that correspond to trips taken over the course of 8 years. There are 113,804 rows of data in the newdata data,frame. str(newdata) 'data.frame': 113804 obs. of 2 variables: $ Season: Factor w/ 4 levels 1,2,3,4: 1 1 1 1 1 1 1 1 1 1 ... $ Area : Factor w/ 3 levels 625,631,Bay: 3 3 3 3 3 3 3 3 3 3 ... Example: Season Area 1 1 Bay 2 1 625 3 1 631 4 2 Bay 5 2 625 6 2 631 7 3 Bay 8 3 625 9 3 631 10 4 Bay 11 4 625 12 4 631 1. Do I need to create dummy variables for all levels for the two variables for input into predict, or does predict function act like zeroinfl where if the variables are coded as factors, this is done automatically by R. 2. Predict returns values of 0.0461 - 0.6015. If I am trying to predict the number of animals how do I interpret this? Since no predicted values are greater than 1 and I need whole numbers, I rounded the predicted data so that any value less than 0.5 was equal to 0 and any value greater than 0.5 was equal to 1. Does this seem correct? Thanks for any help. Sally Roman Fisheries Management Specialist Virginia Marine Resources Commission 2600 Washington Avenue, 3rd Floor Newport News, VA 23607 Phone: 757-247-2243 [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Gregory (Greg) L. Snow Ph.D. 538...@gmail.com __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] posting for R-help on the predict function and categorical variables in R
Hi - I would like to post to the R-help mailing list. Here is my post: -This is more of a general question about how the predict function treats categorical variables and how to interpret the output from predict. I have a zeroinfl model to predict the number of animals encountered: b9-zeroinfl(Count ~ as.factor (Area) + as.factor(Season)|1, dist=negbin,data = total) where Count is the number of animals and the explanatory variables are Area and Season, both are coded as factors in the model. Area has three levels and Season has 4 levels. Coding the two as factors allows for R to create dummy variables for each variable for use in the model. When I use the predict function to predict the number of animals for a larger data set, I want to make sure I'm understanding what is happening. My newdata for predict is: newdata- as.data.frame(Season, Area) Both variables are coded as factors and the dataframe is in a long format. There are records for each combination of Season and Area that correspond to trips taken over the course of 8 years. There are 113,804 rows of data in the newdata data,frame. str(newdata) 'data.frame': 113804 obs. of 2 variables: $ Season: Factor w/ 4 levels 1,2,3,4: 1 1 1 1 1 1 1 1 1 1 ... $ Area : Factor w/ 3 levels 625,631,Bay: 3 3 3 3 3 3 3 3 3 3 ... Example: Season Area 1 1 Bay 2 1 625 3 1 631 4 2 Bay 5 2 625 6 2 631 7 3 Bay 8 3 625 9 3 631 10 4 Bay 11 4 625 12 4 631 1. Do I need to create dummy variables for all levels for the two variables for input into predict, or does predict function act like zeroinfl where if the variables are coded as factors, this is done automatically by R. 2. Predict returns values of 0.0461 - 0.6015. If I am trying to predict the number of animals how do I interpret this? Since no predicted values are greater than 1 and I need whole numbers, I rounded the predicted data so that any value less than 0.5 was equal to 0 and any value greater than 0.5 was equal to 1. Does this seem correct? Thanks for any help. Sally Roman Fisheries Management Specialist Virginia Marine Resources Commission 2600 Washington Avenue, 3rd Floor Newport News, VA 23607 Phone: 757-247-2243 [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] - help with the predict function
Hi all, I would like to predict some values for an nls regression function (functional response model Rogers type II). This is an asymptotic function of which I would like to predict the asymptotic value I estimated the paramters with nls, but can't seem to get predictions for values of m choice.. This is my script: RogersII_N - nls(FR~N0-lambertW(attackR3_N*Th3_N*N0*exp(-attackR3_N*(24-Th3_N*N0)))/(attackR3_N*Th3_N),start=list(attackR3_N=0.04,Th3_N=1.46),control=list(maxiter=1)) lista - c(1,2,100,1000) predict(RogersII_N,newdata=lista) I created a list (lista) with some values of which I would like the predict function to give me function values What am I doing wrong? Kind regards, Met vriendelijke groeten, Joachim Don't waste paper! Think about the environment before printing this e-mail __ Joachim Audenaert Adviesdienst Gewasbescherming Proefcentrum voor Sierteelt Schaessestraat 18 B-9070 Destelbergen Belgium Tel. +32 9 353 94 71 Fax +32 9 353 94 95 E-mail: joachim.audena...@pcsierteelt.be www.pcsierteelt.be __ [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] - help with the predict function
Joachim Audenaert Joachim.Audenaert at pcsierteelt.be writes:
Hi all,
I would like to predict some values for an nls regression function
(functional response model Rogers type II). This is an asymptotic function
of which I would like to predict the asymptotic value
I estimated the paramters with nls, but can't seem to get predictions for
values of m choice..
This is my script:
RogersII_N -
nls(FR~N0-lambertW(attackR3_N*Th3_N*N0*
exp(-attackR3_N*(24-Th3_N*N0)))/(attackR3_N*Th3_N),
start=list(attackR3_N=0.04,Th3_N=1.46),control=list(maxiter=1))
lista - c(1,2,100,1000)
predict(RogersII_N,newdata=lista)
I created a list (lista) with some values of which I would like the
predict function to give me function values
What am I doing wrong?
If you want to use the predict() function, it is best
(necessary?) to use the form of nls() where you pass data
explicitly within a data frame, as the data= argument, instead
of using variables in the global workspace. Also: for this
model, the asymptotic value is (time/(handling time)) = 24/Th3_N ...
so you don't really need predict() (although it is still useful)
library(emdbook)
rogers.pred - function(N0,a,h,T) {
N0 - lambertW(a*h*N0*exp(-a*(T-h*N0)))/(a*h)
}
params0 - list(attackR3_N=0.05,Th3_N=1.5)
params1 - list(attackR3_N=0.04,Th3_N=1.46)
dat - expand.grid(N0=1:20,rep=1:10)
dat$FR0 - with(c(dat,params0),
rogers.pred(N0,a=attackR3_N,h=Th3_N,T=24))
par(las=1,bty=l)
with(dat,plot(N0,FR0,ylim=c(0,12)))
set.seed(101)
dat$FR - rnorm(nrow(dat),dat$FR0,sd=1)
with(dat,points(N0,FR,col=rgb(1,0,0,alpha=0.5)))
RogersII_N - nls(FR~rogers.pred(N0,attackR3_N,Th3_N,T=24),
start=params1,data=dat,control=list(maxiter=1))
lista - c(1,2,100,1000,1)
predict(RogersII_N,newdata=data.frame(N0=lista))
N0vec - 1:20
lines(N0vec,predict(RogersII_N,newdata=data.frame(N0=N0vec)),
col=4)
with(as.list(coef(RogersII_N)),24/Th3_N)
__
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
