outlierCriteria: a function for several outlier criteria

In sanchezi/phisStatR: Phenoarch Greenhouse Data Analysis

Description

this function calculates outlier criteria on plants for each parameter

Usage

outlierCriteria(datain, typeD, residin, typeR, trait, resRawName,
  resStdName, threshold)

Arguments

datain	input dataframe of parameters
typeD	type of datain dataframe (1==wide, 2==long)
residin	input dataframe of residuals
typeR	type of residin dataframe (1==wide, 2==long)
trait	character, trait of interest to model (example biovolume24, PH24 ...)
resRawName	character, names of the raw residual in datain
resStdName	character, names of the standardized residual in datain
threshold,	numeric threshold for the normal quantile in raw criteria

Details

This function needs in input a dataframe with residuals extracted from a mixed linear model (for instance using asreml or nlme libraries) and an another dataframe with the estimated parameters (biovolume, plantHeight, leafArea etc...). Several criteria will be calculated using different types of residuals. The 2 input dataframe must contain the following columns names: "experimentAlias","Line" and "Position".

raw and quartile criteria

use raw residuals

influence criterion

uses standardized residuals

The function must be executed for each parameter of interest: biovolume, plantHeight and phy. Each criteria will be used according with some rules:

Small plant

biovolume and phy

Big plants

biovolume and plantHeight

Value

a dataframe with columns identifiying criteria used to detect outlier plants with 1==plant OK - 0==plant KO to suppress

critraw

raw criterion, critci: quartiles criterion

critinfl

influence criterion with standardized residuals

Examples

# Not run
# dt1<-outlierCriteria()

sanchezi/phisStatR documentation built on Nov. 14, 2019, 7:10 p.m.