IPV 1.0.0
We thank an anonymous reviewer for suggesting many of the
improvements in this version!
Added features
- Introduced class "IPV"for the output of the estimation
functionipv_estand added first version of functionsis.IPV,validate_IPV, andnew_IPV.
- Chart functions now all work (exclusively) on objects of class
"IPV"with facet_chart and item_chart having the user
specify which of the tests should be plotted via thetestargument in nested cases.
- ipv_estand input functions now (exclusively) produce
output of class- "IPV".
- Estimation function ipv_estcan now pass further
arguments to the underlying lavaan estimation functionlavaan::cfa. Among other things, this makes it possible to
choose the estimator and treat missing values in various ways.
- Changed example dataset HEXACOto also include cases
with missing values.
- ipv_estnow also works for data in long format.
- Example on how to add ggplot layers to an IPV chart to the vignette
(this has been possible before, but should be emphasized)
- Optional markers for the center distances in facet and nested charts
emphasize that the values are read at the border of the circles
(included by default).
- When creating an item_overview, it is now possible to
scale the font size (size_font =), specify the number of
rows per facet to wrap overly wide charts (wrap =), scale
the width (width =) and height (height =) of
the file output, and select a subset of tests (tests =) and
facets (facets =) for display.
Bugfixes
- resolved a warning message for nested charts that include tests
without facets (no effect on output)
- Function renamerenamed torelabelto
avoid conflicts with other packages.
- Resolved multiple issues with inconsistent positioning of labels in
item_overviewcharts and changed the default font to
‘sans’, since the consistent width of monospaced letters is no longer
needed.
- Changed several examples to include more appropriate models at the
expense of runtime.
IPV 0.2.0
Added Features
- Package now works based on raw data with automated model estimation
using lavaan (function ipv_est). This is now the
recommended workflow!
- New raw example data (HEXACO) from the open psychometrics project
(see?HEXACO).
- Create an overview over all (squared) factor loadings in all three
models on which a nested chart is based (function
item_overview).
- Refined method available to compute center distances for groups of
items (parameter cd_method = "aggregate"). This is now the
default method!
- Show only a section of a chart (parameters zoom_xandzoom_y)
- Convenient changing of test, facet, or item labels (function
rename).
- Rotate the tick label in facet charts and nested charts to avoid
overlap (parameter rotate_tick_label).
- Set the limit of the grid in item charts (parameter
grid_limit).
- Set the order of facets for all charts types (parameter
facet_order).
- Set the order of tests for nested charts (parameter
test_order).
- Add an overall title for all chart types (parameter
title).
- Draw xarrows in a nested chart that end on a test circle (instead of
a facet circle) if facet1orfacet2is
specified asNAfor that arrow.
- Overhaul of the documentation in the vignette.
Bugfixes
- Standardized the default order of facets across facet charts and
item charts. It is now consistently the order of the correlation matrix
columns in data =.
- Negative correlations between facets or tests should now be
displayed correctly.
- Removed an erroneous message claiming that the axis tick was set
automatically when it was set manually.
- The axis tick mark (gray dotted circle) now disappears “behind”
circles of facet charts and nested charts, instead of cutting through
them.
- the axis tick label (number) now more reliably appears close to the
axis tick mark in facet charts and nested charts.
- removed the argument show_xarrowsinnested_chart, which only existed for technical
reasons.
IPV 0.1.0
Release
IPV 0.1.1.