| Type: | Package |
| Title: | Person-Centered Analysis |
| Version: | 0.2.1 |
| Maintainer: | Joshua M Rosenberg <jmichaelrosenberg@gmail.com> |
| Description: | Provides an easy-to-use yet adaptable set of tools to conduct person-center analysis using a two-step clustering procedure. As described in Bergman and El-Khouri (1999) <doi:10.1002/(SICI)1521-4036(199910)41:6%3C753::AID-BIMJ753%3E3.0.CO;2-K>, hierarchical clustering is performed to determine the initial partition for the subsequent k-means clustering procedure. |
| License: | MIT + file LICENSE |
| URL: | https://github.com/jrosen48/prcr |
| BugReports: | https://github.com/jrosen48/prcr/issues |
| LazyData: | TRUE |
| Imports: | dplyr, tidyr, ggplot2, tibble, irr, lpSolve, purrr, class, forcats, magrittr |
| Suggests: | rmarkdown, knitr, devtools |
| VignetteBuilder: | knitr |
| RoxygenNote: | 7.0.2 |
| Depends: | R (≥ 2.10) |
| NeedsCompilation: | no |
| Packaged: | 2020-02-09 13:08:49 UTC; jrosenb8 |
| Author: | Joshua M Rosenberg [aut, cre], Jennifer A Schmidt [aut], Patrick N Beymer [aut], Rebecca R Steingut [ctb] |
| Repository: | CRAN |
| Date/Publication: | 2020-02-09 17:00:05 UTC |
Create profiles of observed variables using two-step cluster analysis
Description
Create profiles of observed variables using two-step cluster analysis
Usage
create_profiles_cluster(
df,
...,
n_profiles,
to_center = FALSE,
to_scale = FALSE,
distance_metric = "squared_euclidean",
linkage = "complete"
)
Arguments
df |
with two or more columns with continuous variables |
... |
unquoted variable names separated by commas |
n_profiles |
The specified number of profiles to be found for the clustering solution |
to_center |
Boolean (TRUE or FALSE) for whether to center the raw data with M = 0 |
to_scale |
Boolean (TRUE or FALSE) for whether to scale the raw data with SD = 1 |
distance_metric |
Distance metric to use for hierarchical clustering; "squared_euclidean" is default but more options are available (see ?hclust) |
linkage |
Linkage method to use for hierarchical clustering; "complete" is default but more options are available (see ?dist) |
Details
Function to create a specified number of profiles of observed variables using a two-step (hierarchical and k-means) cluster analysis.
Value
A list containing the prepared data, the output from the hierarchical and k-means cluster analysis, the r-squared value, raw clustered data, processed clustered data of cluster centroids, and a ggplot object.
Examples
d <- pisaUSA15
m3 <- create_profiles_cluster(d,
broad_interest, enjoyment, instrumental_mot, self_efficacy,
n_profiles = 3)
summary(m3)
Identifies potential outliers
Description
Identifies potential outliers
Usage
detect_outliers(df, return_index = TRUE)
Arguments
df |
data.frame (or tibble) with variables to be clustered; all variables must be complete cases |
return_index |
Boolean (TRUE or FALSE) for whether to return only the row indices of the possible multivariate outliers; if FALSE, then all of the output from the function (including the indices) is returned |
Details
* add an argument to ‘create_profiles_cluster()' to remove multivariate outliers based on Hadi’s (1994) procedure
Value
either the row indices of possible multivariate outliers or all of the output from the function, depending on the value of return_index
Estimates R^2 (r-squared) values for a range of number of profiles
Description
Estimates R^2 (r-squared) values for a range of number of profiles
Usage
estimate_r_squared(
df,
...,
to_center = FALSE,
to_scale = FALSE,
distance_metric = "squared_euclidean",
linkage = "complete",
lower_bound = 2,
upper_bound = 9,
r_squared_table = TRUE
)
Arguments
df |
with two or more columns with continuous variables |
... |
unquoted variable names separated by commas |
to_center |
(TRUE or FALSE) for whether to center the raw data with M = 0 |
to_scale |
Boolean (TRUE or FALSE) for whether to scale the raw data with SD = 1 |
distance_metric |
Distance metric to use for hierarchical clustering; "squared_euclidean" is default but more options are available (see ?hclust) |
linkage |
Linkage method to use for hierarchical clustering; "complete" is default but more options are available (see ?dist) |
lower_bound |
the smallest number of profiles in the range of number of profiles to explore; defaults to 2 |
upper_bound |
the largest number of profiles in the range of number of profiles to explore; defaults to 9 |
r_squared_table |
if TRUE (default), then a table, rather than a plot, is returned; defaults to FALSE |
Details
Returns ggplot2 plot of cluster centroids
Value
A list containing a ggplot2 object and a tibble for the R^2 values
student questionnaire data with four variables from the 2015 PISA for students in the United States
Description
student questionnaire data with four variables from the 2015 PISA for students in the United States
Usage
pisaUSA15
Format
Data frame with columns #'
- CNTSTUID
international student ID
- SCHID
international school ID
...
Source
http://www.oecd.org/pisa/data/
Return plot of profile centroids
Description
Return plot of profile centroids
Usage
plot_profiles(d, to_center = F, to_scale = F)
Arguments
d |
summary data.frame output from create_profiles_cluster() |
to_center |
whether to center the data before plotting |
to_scale |
whether to scale the data before plotting |
Details
Returns ggplot2 plot of cluster centroids
Value
A ggplot2 object
Prints details of prcr cluster solution
Description
Prints details of prcr cluster solution
Usage
## S3 method for class 'prcr'
print(x, ...)
Arguments
x |
A 'prcr' object |
... |
Additional arguments |
Details
Prints details of of prcr cluster solution
Concise summary of prcr cluster solution
Description
Concise summary of prcr cluster solution
Usage
## S3 method for class 'prcr'
summary(object, ...)
Arguments
object |
A 'prcr' object |
... |
Additional arguments |
Details
Prints a concise summary of prcr cluster solution