umxEFA: umxEFA

Description

Perform full-information maximum-likelihood factor analysis on a data matrix. as in factanal, you need only specify the number of factors and offer up some manifest data, e.g:

umxEFA(factors = 2, data = mtcars)

Usage

umxEFA(x = NULL, factors = NULL, data = NULL, covmat = NULL,
  n.obs = NULL, scores = c("none", "ML", "WeightedML", "Regression"),
  minManifests = NA, rotation = c("varimax", "promax", "none"),
  name = "efa", digits = 2, report = c("markdown", "html"))

Arguments

Either 1: data, 2: A formula (not implemented yet), 3: A vector of variable names, or 4: A name for the model.

factors

Either number of factors to request or a vector of factor names.

data

A dataframe of manifest columns you are modeling

covmat

Covariance matrix of data you are modeling (not implemented)

n.obs

Number of observations in covmat (if provided, default = NA)

scores

Type of scores to produce, if any. The default is none, "Regression" gives Thompson's scores. Other options are 'ML', 'WeightedML', Partial matching allows these names to be abbreviated.

minManifests

The least number of variables required to return a score for a participant (Default = NA).

rotation

A rotation to perform on the loadings (default = "varimax" (orthogonal))

name

A name for your model

digits

rounding (default = 2)

report

Report as markdown to the console, or open a table in browser ("html")

Value

- EFA mxModel

Details

Equivalently, you can also give a list of factor names:

umxEFA(factors = c("g", "v"), data = mtcars)

The factor model is implemented as a structural equation model, e.g.

You can request scores from the model. Unlike factanal, these can cope with missing data.

In an EFA, all items may load on all factors. For identification we need m^2 degrees of freedom. We get m * (m+1)/2 from fixing factor variances to 1 and covariances to 0. We get another m(m-1)/2 degrees of freedom by fixing the upper-right hand corner of the factor loadings component of the A matrix. The manifest variances are also lbounded at 0.

EFA reports standardized loadings: to do this, we scale the data.

Bear in mind that factor scores are indeterminate and can be rotated.

This is very much early days.

References

- http://github.com/tbates/umx

Examples

Run this code

# NOT RUN {
myVars <- c("mpg", "disp", "hp", "wt", "qsec")
m1 = umxEFA(mtcars[, myVars], factors =   2, rotation = "promax")
loadings(m1)
m2 = factanal(~ mpg + disp + hp + wt + qsec, factors = 2, rotation = "promax", data = mtcars)
loadings(m2)
# }
# NOT RUN {
plot(m2)
m1 = umxEFA(myVars, factors = 2, data = mtcars, rotation = "promax")
m1 = umxEFA(name = "named", factors = "g", data = mtcars[, myVars])
m1 = umxEFA(name = "by_number", factors = 2, rotation = "promax", data = mtcars[, myVars])
x = umxEFA(name = "score", factors = "g", data = mtcars[, myVars], scores= "Regression")
# }

Run the code above in your browser using DataCamp Workspace