Flexible statistical test

interplot 0.2.3

New features

1. Adding new statistics for testing the statistical significance of the conditional effect (by stats_cp).
2. Adding an argument to modify the plot caption (by txt_caption).

Bugfix

1. Adding examples of models based on multiple imputations.
2. Modifying the vignette to include new functions.

`facet_labs`

Function Updates

• Adds facet_labs argument, an optional character vector of facet labels to be used when plotting an interaction with a factor variable.
• interplot no longer sets the random seed; to ensure complete reproducibility, users must now set their own seeds using set.seed() before calling interplot functions.

Diff (Max - Min) CI

Function Updates

• Showing the confidence intervals between the conditional effects at the minimum and maximum values of the conditioning variable.

Error Fixed

• Avoiding the warning caused by the class(m) == "polr".

Flexible CI

Function Improvement

• Adding an argument to adjust the CIs.

Bug Fixed for `lmer`

Error Fixed

• Fixed the error in plotting lmer projects.

Histogram Improvement

Function Updates

• Take the steps argument back in case of special design requirement of the plot.
• Fixed an error in presenting the histogram on categorical conditioning variables.
• Improving the histogram presentation: all the bars for categorical variables are centered.

Vignette Updates

Updated the vignette including instructions of how to change the aesthetics of the plot and how to use histogram function.

Histogram function

A function to draw a histogram under the conditional effect plot is available. The aesthetics of the components in the conditional effect plot are manageable now.

Error Fixed

Two errors are fixed in this version:

1. The function only present the substantive coefficients of multilevel model results excluding the variances matrix.
2. The presentation on quadratic terms is corrected (Thanks for Brenton's contribution!).

Factor base term and ggplot2 2.0

The version adds functions to deal with interactions with one factor base term, and the whole package works for ggplot2 2.0.0 now.

Fixed bug for nonlinear mlm and glm

fix the bug for nonlinear multilevel models with multiply imputed data (gmlmmi).