Package 'dualScale'

Title: Dual Scaling Analysis of Data
Description: Dual Scaling, developed by Professor Shizuhiko Nishisato (1994, ISBN: 0-9691785-3-6), is a fundamental technique in multivariate analysis used for data scaling and correspondence analysis. Its utility lies in its ability to represent multidimensional data in a lower-dimensional space, making it easier to visualize and understand underlying patterns in complex data. This technique has been implemented to handle various types of data, including Contingency and Frequency data (CF), Multiple-Choice data (MC), Sorting data (SO), Paired-Comparison data (PC), and Rank-Order data (RO), providing users with a powerful tool to explore relationships between variables and observations in various fields, from sociology to ecology, enabling deeper and more efficient analysis of multivariate datasets.
Authors: Jose G. Clavel [aut] , Shizuiko Nishisato [aut] , Roberto de la Banda [aut, cre] , Antonio Pita [ctb]
Maintainer: Roberto de la Banda <[email protected]>
License: AGPL (>= 3)
Version: 1.0.0
Built: 2024-12-03 06:41:34 UTC
Source: CRAN

Help Index


Nishisato and Clavel, artificial set of data

Description

10 observation and 3 variables erroneously coded.

Usage

bad_coded

Format

A data set with 10 observations on the following 3 variables:

V2

Option 1 is omited

V3

Options go from 1 to 8

V4

Option 1 omited, other are not consecutive and there is NA

Details

The data were collected from 23 participants at a workshop in Singapore in 1985

Source

Nishisato, S. and Baba, Y. (1999). On contingency, projection and forced classification of dual scaling. Behaviormetrika, 26, 207–219.

References

Nishisato, S. (2007). Multidimensional Nonlinear Descriptive Analysis. Chapman & Hall/CRC.


Christmas party plans

Description

As a course assignment for Nishisato's class, Ian Wiggins, a student, collected paired comparison data from 14 researchers at a research institute on his eight Christmas party plans.

Usage

christmas

Format

A subset of the original data set of 14 subjects on 15 pairs of 6 plans:

plan1

A pub/restaurant crawl after work

plan2

A reasonably priced lunch in an area restaurant

plan3

Keep to one's self

plan4

An evening banquet at a restaurant

plan5

A pot-luck at someone's home after work

plan6

A ritzy lunch at a good restaurant (tablecloths)

Details

The data were originally collected from 14 participants by 28 pairs of plans with elements 1 for the choice of the first plan and 2 for the choice of the second plan. For computations, 2 is converted to -1.

Source

Nishisato, S. and Nishisato, I.(1994). Dual Scaling in a Nutshell. Toronto: MicroStats.

References

Nishisato, S. (2022). Optimal Quantification and Symmetry. Behaviormetrika, 12, 137.


Curricula and Social classes

Description

Hollingshead (1949) found that the members of a small Middle Western community in the United States divided themselves into 5 social classes. He investigated his prediction that adolescents in the different social classes would enroll in different curricula

Usage

curricula

Format

A data set of 390 subjects on 5 social classes and 3 different curricula:

s.class1

Merged social classes I and II because the frequencies were small

s.class2

Social class III

s.class3

Social class IV

s.class4

Social class V

curricula1

College Prep curriculum

curricula2

General curriculum

curricula3

Commercial curriculum

Details

The data were originally collected from 390 participants in terms of their social classes and actual curriculum enrollments.

Source

Nishisato, S. and Nishisato, I.(1994). Dual Scaling in a Nutshell. Toronto: MicroStats.

References

Hollingshead, A.B. (1949). Elmtown's Youth: The Impact of Social Classes on Adolescents. Wiley.


Contingency and frequency data analysis

Description

Contingency and frequency data analysis

Usage

ds_cf(input, solutions = NULL)

Arguments

input

A data set with valid data

solutions

Optional arguments. A number of intended solutions

Value

call

Call with all of the specified arguments are specified by their full names

orig_data

Initial data

item_op_lbl

Item options labels

sub_lbl

Subjects options labels

tot_row

Sum of subject values

tot_row

Sum of item values

solutions

Maximum possible solutions

out

Results obtained

norm_opt

Normed option weights

proj_opt

Projected option weights

norm_sub

Normed subject scores

proj_sub

Projected subject scores

appro0

Order 0 approximation for initial data

approx

Order k approximation for each solution

residual0

Residual matrix for initial data

residual

Residual matrix k for each solution

Examples

ds_cf(curricula)
ds_cf(preferences)

Multiple choice data analysis

Description

Multiple choice data analysis

Usage

ds_mc(input, solutions = NULL, mode = c("rad", "act"))

Arguments

input

A data set with valid data

solutions

Optional argument. A number of intended solutions

mode

Optional argument. In case of NA values, the action to be taken. See help("ds_mc_check") for more information. Radical action by default.

Value

call

Call with all of the specified arguments are specified by their full names

orig_data

Initial data

item_op_lbl

Item options labels

sub_lbl

Subjects options labels

solutions

Maximum possible solutions

out

Results obtained

item_stat

Item statistics

info

Distribution of component

rij

Inter item correlation

proj_opt

Projected option weights

proj_sub

Projected subject scores

norm_opt

Normed option weights

norm_sub

Normed subject scores

See Also

ds_mc_check()

Examples

ds_mc(singaporean)
ds_mc(singaporean, solutions = 2)

Function to identify incorrect Multiple Choice input data

Description

Function to identify incorrect Multiple Choice input data

Usage

ds_mc_check(input, mode = c("rad", "act"))

Arguments

input

The input data to be checked

mode

Do you want to use a radical ("rad") correction mode or active ("act") allocations?

Value

A list with the original input and the converted input

Examples

ds_mc_check(singaporean)
ds_mc_check(bad_coded)

Forced multiple choice data analysis

Description

Forced multiple choice data analysis

Usage

ds_mcf(input, crit, solutions = NULL, mode = c("rad", "act"))

Arguments

input

A data set with valid data

crit

Used to determine a criterion item for forced classification analysis

solutions

Optional argument. A number of intended solutions

mode

Correction mode to incorrect data.

Details

There are three types of outputs: Forced classification of the criterion item (type A); dual scaling of non-criterion items by ignoring the criterion item (type B); dual scaling of non-criterion items after eliminating the influence of the criterion item (type C). These three types correspond to, respectively, dual scaling of data projected onto the subspace of the criterion item, dual scaling of non-criterion items, and dual scaling of data in the complementary space of the criterion item.

Value

call

Call with all of the specified arguments are specified by their full names

orig_data

Initial data

crit_item

The criterion item for forced classification

item_op_lbl

Item options labels

sub_lbl

Subjects options labels

solutions_mcf

Maximum possible solutions for forced multiple choice

solutions_mc

Maximum possible solutions for multiple choice

info_\emph{x}

Distribution of component information according to output

out_\emph{x}

Results obtained according to output

item_stat_\emph{x}

Item statistics according to output (Not type C)

rij_\emph{x}

Inter item correlation according to output (Not type C)

proj_opt_\emph{x}

Projected option weights according to output

proj_sub_\emph{x}

Projected subject scores according to output

norm_opt_\emph{x}

Normed option weights according to output

norm_sub_\emph{x}

Normed subject scores according to output

match_missmatch

Match-mismatch tables

predict

Percentage of correct classification

comp_cont

Component contamination

tot_cont

Total contamination

See Also

ds_mc_check()

Examples

ds_mcf(singaporean, crit = 1)

Paired comparison data analysis

Description

Paired comparison data analysis

Usage

ds_pc(input, solutions = NULL)

Arguments

input

A data set with valid data

solutions

Optional argument. A number of intended solutions

Value

call

Call with all of the specified arguments are specified by their full names

orig_data

Initial data

item_op_lbl

Item options labels

sub_lbl

Subjects options labels

solutions

Maximum possible solutions

out

Results obtained

mat_e

Matrix E

norm_opt

Normed option weights

proj_opt

Projected option weights

norm_sub

Normed subject scores

proj_sub

Projected subject scores

Examples

ds_pc(christmas)

Rank order data analysis

Description

Rank order data analysis

Usage

ds_ro(input, solutions = NULL)

Arguments

input

A data set with valid data

solutions

Optional argument. A number of intended solutions

Value

call

Call with all of the specified arguments are specified by their full names

orig_data

Initial data

item_op_lbl

Item options labels

sub_lbl

Subjects options labels

solutions

Maximum possible solutions

out

Results obtained

mat_e

Matrix E

norm_opt

Normed option weights

proj_opt

Projected option weights

norm_sub

Normed subject scores

proj_sub

Projected subject scores

out_rank

Results obtained by rank analysis

norm_opt_rank

Normed option weights by rank analysis

proj_opt_rank

Projected option weights by rank analysis

norm_rank

Normed rank scores

proj_rank

Projected rank scores

Examples

ds_ro(goverment)

Goverment services and facilities

Description

A data set collected in Nishisato's scaling class (1982) in which 31 students on 10 municipal services in Toronto.

Usage

goverment

Format

A subset of the original data of 10 subjects on 10 municipal services in Toronto:

A

Public transit system

B

Postal service

C

Medical care, including hospitals and clinics

D

Sports, recreational facilities

E

Police protection

F

public libraries

G

cleaning streets

H

restaurants

I

theatres

J

Overall planning and development

Details

The data were originally collected to ranked the "most satisfactory" service, the "second most satisfactory", and so on until the "least satisfactory".

Source

Nishisato, S. and Nishisato, I.(1994). Dual Scaling in a Nutshell. Toronto: MicroStats.

References

Nishisato, S. and Nishisato, I.(1994). Dual Scaling in a Nutshell. Toronto: MicroStats.


Obtain the data used in the graphs

Description

Obtain the data used in the graphs

Usage

plot_data(x, dim1 = 1, dim2 = 2, type = c("Asy1", "Asy2", "Sub", "Ite"), ...)

Arguments

x

A Dual Scale object

dim1

Component for the horizontal axis. Default dimension 1

dim2

Component for the vertical axis. Default dimension 2

type

Graph type

Asy1

Assymetric graph for projected options versus normed subjects (default)

Asy2

Assymetric graph for normed options versus projected subjects

Sub

Only subjects graph

Ite

Only items graph

...

Arguments to be passed to methods

Value

A dataframe with the data used

Examples

plot_data(ds_cf(curricula))
plot_data(ds_mc(singaporean))
plot_data(ds_mcf(singaporean, crit = 1))
plot_data(ds_pc(christmas))
plot_data(ds_ro(goverment))

Plot of Dual Scale analysis

Description

Plot of Dual Scale analysis

Usage

## S3 method for class 'dualScale'
plot(x, dim1 = 1, dim2 = 2, type = c("Asy1", "Asy2", "Sub", "Ite"), ...)

Arguments

x

A Dual Scale object

dim1

Component for the horizontal axis. Default dimension 1

dim2

Component for the vertical axis. Default dimension 2

type

Graph type

Asy1

Assymetric graph for projected options versus normed subjects (default)

Asy2

Assymetric graph for normed options versus projected subjects

Sub

Only subjects graph

Ite

Only items graph

...

Arguments to be passed to methods

Value

A plot of the available information from the object

See Also

plot(),ggplot2::ggplot2()

Examples

plot(ds_cf(curricula))
plot(ds_mc(singaporean))
plot(ds_mcf(singaporean, crit = 1))
plot(ds_pc(christmas))
plot(ds_ro(goverment))

Preferences, artificial set of data

Description

Artificial set of data where 13 people were asked two questions.

Usage

preferences

Format

A data set of contingency responses:

A, B, C

Do you prefer coffee to tea? (Yes, Not always, No)

Y, N

Do you smoke? (Yes, No)

Details

Artificial set of data where 13 people were asked two questions.

Source

Nishisato, S. (1980). Analysis of categorical data: Dual Scaling and its Applications. University of Toronto: Heritage.

References

Nishisato, S. (1980). Analysis of categorical data: Dual Scaling and its Applications. University of Toronto: Heritage.


Print of Dual Scale analysis

Description

print method for package "dualScale"

Usage

## S3 method for class 'dualScale'
print(x, ...)

Arguments

x

An dualScale object for which a summary is desired

...

Arguments to be passed to methods

Value

A print of the available information from the object

See Also

print()

Examples

print(ds_cf(curricula))
print(ds_cf(preferences))
print(ds_mc(singaporean))
print(ds_mcf(singaporean, crit = 1))
print(ds_pc(christmas))
print(ds_ro(goverment))

Singaporean children as viewed by adults

Description

A short survey on childrem in Singapore.

Usage

singaporean

Format

A data set of 23 subjects on 4 multiple-choice questions:

A

How old are you? (1 = 20-29, 2 = 30-39, 3 = 40 or over)

B

Children today are not as disciplined as when I was a child (1 = agree, 2 = disagree, 3 = I cannot tell)

C

Children today are not as fortunate as when I was a child (1 = agree, 2 = disagree, 3 = I cannot tell)

D

Religions should be taught at school (1 = agree, 2 = disagree, 3 = Indifferent)

Details

The data were collected from 23 participants at a workshop in Singapore in 1985

Source

Nishisato, S. and Nishisato, I.(1994). Dual Scaling in a Nutshell. Toronto: MicroStats.

References

Nishisato, S. (2007). Multidimensional Nonlinear Descriptive Analysis. Chapman & Hall/CRC.


Summary of Dual Scale analysis

Description

summary method for class "dualScale"

Usage

## S3 method for class 'dualScale'
summary(object, ...)

Arguments

object

An dualScale object for which a summary is desired

...

Arguments to be passed to methods

Value

A summary of the available information from the object

See Also

summary()

Examples

summary(ds_cf(curricula))
summary(ds_cf(preferences))
summary(ds_mc(singaporean))
summary(ds_mcf(singaporean, crit = 1))
summary(ds_pc(christmas))
summary(ds_ro(goverment))