epr (version 3.0)

pr2: Analysis of polynomial regression

Description

The function performs analysis of polynomial regression in simple designs with quantitative treatments. This function performs analysis the lack of fit .

Usage

pr2(data, design = 1, list = FALSE, type = 2)

Arguments

data

data is a data.frame

data frame with two columns, treatments and response (completely randomized design)

data frame with three columns, treatments, blocks and response (randomized block design)

data frame with four columns, treatments, rows, cols and response (latin square design)

data frame with five columns, treatments, square, rows, cols and response (several latin squares)

design

1 = completely randomized design

2 = randomized block design

3 = latin square design

4 = several latin squares

list

FALSE = a single response variable

TRUE = multivariable response

type

type is form of obtain sum of squares

1 = a sequential sum of squares

2 = a partial sum of squares

Value

Returns analysis of variance, models, t test for coefficients and R squared and adjusted R squared.

Details

The response and the treatments must be numeric. Other variables can be numeric or factors.

References

KAPS, M. and LAMBERSON, W. R. Biostatistics for Animal Science: an introductory text. 2nd Edition. CABI Publishing, Wallingford, Oxfordshire, UK, 2009. 504p.

SAMPAIO, I. B. M. Estatistica aplicada a experimentacao animal. 3nd Edition. Belo Horizonte: Editora FEPMVZ, Fundacao de Ensino e Pesquisa em Medicina Veterinaria e Zootecnia, 2010. 264p.

See Also

lm, lme(package nlme), ea1(package easyanova), pr1, regplot

Examples

# NOT RUN {
# analysis in completely randomized design
data(data1)
r1=pr2(data1)
names(r1)
r1
r1[1]

# analysis in randomized block design
data(data2)
r2=pr2(data2, design=2)
r2

# analysis in latin square design
data(data3)
r3=pr2(data3, design=3)
r3

# analysis in several latin squares
data(data4)
r4=pr2(data4, design=4)
r4

# data
treatments=rep(c(0.5,1,1.5,2,2.5,3), c(3,3,3,3,3,3))
r1=rnorm(18,60,3)
r2=r1*1:18
r3=r1*18:1
r4=r1*c(c(1:10),10,10,10,10,10,10,10,10)
data6=data.frame(treatments,r1,r2,r3, r4)

# use the argument list = TRUE
pr2(data6, design=1, list=TRUE)


# }