Skip to content

A Collection of Utilities for the Advanced Workshop in Sensory Evaluation of SPISE 2018.

Notifications You must be signed in to change notification settings

HerveAbdi/R4SPISE2018

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

26 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

title author date
Package: R4SPISE2018
Hervé Abdi
`October 28, 2018`

R4SPISE2018 .0.1.0

A set of data sets, scripts, and vignettes used for the Advanced Workshop of the 2018 SPISE Meeting that took place in Da Nang, Vietnam: July 20 to July 30, 2018.

Workshop: individual differences in sensory evaluation

Packages to install prior to installing this package

Prior to installing the R4SPISE2018 package, we need to install some other packages (mostly from Github).

# decomment the line below if devtools is not yet installed
# install.packages('devtools') 
install.packages('factoextra')
devtools::install_github('HerveAbdi/PTCA4CATA')
devtools::install_github('HerveAbdi/DistatisR')
devtools::install_github('HerveAbdi/data4PCCAR')

How to install the package with devtools < 2.0

To install the package R4SPISE2018, with devtools older than version 2.0, use the following R-command:

# install.packages('devtools') 
# decomment this line if devtools is not yet installed
devtools::install_github('HerveAbdi/R4SPISE2018', 
       dependencies = TRUE, # use it first time only, comment after
       build_vignettes = TRUE) # to get the vignettes       

The installation of the "dependent" packages can take some time and building the vignettes can take also a bit of time to build: So, be patient!

How to install the package with devtools > 2.0

IMPORTANT: From version 2.0 to 2.1, devtools has changed its options and now the parameter build_vignettes is (silently) ignored. This option is now integrated in the parameter build_opts which passes options for CMD build and which, by default include the following options build_opts = c("--no-resave-data", "--no-manual", "--no-build-vignettes"). However, versions of devtools > 2.2 seems to have resinstated the build_vignettes option. So with devtools version 2.0 or more recent, the option build_opts = c("--no-resave-data", "--no-manual", "--no-build-vignettes") needs to be eliminated in the new call to devtools and replaced by the new option build_opts as:

# install.packages('devtools') 
# decomment the line above if devtools is not yet installed 
install_github("HerveAbdi/R4SPISE2018",
dependencies = TRUE, # use it first time only, comment after
build_opts = c("--no-resave-data", "--no-manual"), # Explicitely get the option
build_vignettes = TRUE) # to get the vignettes  

The installation of the "dependent" packages can take some time and building the vignettes can take also a bit of time to build: So, be patient!

How to build the vignettes

If you have forgotten to build the vignettes, you can build them with the command:

devtools::build_vignettes()

Again, be prepared to wait till the vignettes are built.

Main statistical techniques and R-packages used

  1. Distatis: package DistatisR
  2. Partial Triadic Correspondence Analysis (PTCA): package PTCA4CATA
  3. Various helper functions for Multiple Correspondence Analysis from package data4PCCAR
  4. Multiple Correspondence Analysis (MCA): package ExPosition

Where to download the packages

  1. ExPosition can be downloaded from CRAN
  2. DistatisR, PTCA4CATA, and data4PCCARshould be downloaded from HerveAbdi/Github.

Current Vignettes

beersCATA

Correspondence Analysis for CATA:

Novices and experts evaluated 9 beers.

cheeseMCA

Multiple Correspondence Analysis:

Analyzing a survey about cheese

fermentationIn2NationsMCA

MCA with supplementary observations and variables; and new graphs.:

Analysis of a survey about attitudes towards fermented food answered by 373 participants (220 French and 183 Vietnamese). The demographics of the participants is used as supplementary variables. In addition, 30 participants from the SPISE2018 advanced workshop answered the questionnaire and are considered as supplementary observations.

This example also shows how to create PCA -like graphs for the qualitative variables: correlation, contributions, and bootstrap ratios. It uses several news functions from the package data4PCCAR (see HerveAbdi/data4PCCAR fro, Github )

multiculturalSortingSpices

DISTATIS:

Analyzing a sorting task with different groups of assessors.

sortingNoodlesInAsia

DISTATIS (with vocabulary and barycentric projections):

Analyzing a sorting task of pictures of Ramen Noodles with verbal description of the groups made by the participants.

About

A Collection of Utilities for the Advanced Workshop in Sensory Evaluation of SPISE 2018.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages