Added the argument passes to
eq_models() for checking empirical
equivalences in two or more datasets.
The additional dataset are
resample-with-replacement versions of
the original dataset.
(0.1.0.2)
Added est_summary() to summarize
parameter estimates across models
generated by eq_models(). The output
has a print method (print.est_summary()).
(0.1.0.3, 0.1.0.4)
Added the option to define empirical
equivalence using the implied
covariance matrix (and mean vector, if
available). Whether fit measures or
implied matrices and vectors are used
is determined by the argument
eq_mode. (0.1.0.5)
Added the argument keep_covs to
eq_df_models(). By
default, all covariances will not be
changed because it is unlikely that
users would like to find models that
change a bidirectional path (no
assertion on the causal direction)
to a directional path. This is likely
to be true for covariances among
predictors.
(0.1.0.7)
eq_df_models() in searching
models. Fixed a bug in digest_partable().
No more redundant files -1df models
in the search.
(0.1.0.7)Updated only the GitHub README.md.
No change to the codebase. Identical
to 0.1.0 otherwise.
(0.1.0.1)
Added use_mi to add_k() to use
modification indices to rule out
models unlikely to be equivalent to
the original model. Disabled for now.
(0.1.0.6)
Added exclude_x_y_ecov_interim and
exclude_x_y_ecov_add_drop to allow
exclusion in more stages. Disabled
for now (set to FALSE by default).
(0.1.0.6)
Replaced "searching for" by "generating" in the progress printout. (0.1.0.8)
First public release for testing. (0.0.1)
Added tests using dagitty.
(0.0.1.1)
Updated eq_models_internal()
to automatically mark the original
models, and updated partables_plots()
to retrieve the original model if
marked. (0.0.1.2)
Updated eq_df_models() and
add_k() to handle errors from
modelbpp::gen_models().
(0.0.1.3)
Made the progress bars of
eq_df_models() more informative.
(0.0.1.4, 0.1.0)