dsem 1.7.0

Add option to turn off useless NA/NaN function evaluation function evaluations from nlminb (enabled by default, but overriden in dsem_control)
Use Matrix::bandSparse to simplify logic in make_matrices
Add experimental function partition_variance
Extend total_effect to compute result for either pulse or press experiment
Adding exploratory option gmrf_parameterization = "conditional_krig" that excludes family = "fixed" variables from the GMRF density calculation while still conditioning upon their value(s)

dsem 1.6.0

Export and document make_matrices
Remove ggm from Imports (because it Imports graph which is not on CRAN) and instead define a locally copy of relevant functions. Also adding Giovanni M. Marchetti (the maintainer of ggm as contributor in DESCRIPTION)
Update test_dsep(.) to allow options for imputing missing data prior to test, and using default impute_data = "by_test" based on explorations to date
Add example for total_effect
Adding “model structura” vignette and re-organizing pkgdown page

Add option to simulate data with or without missingness pattern of original data
Change test_dsep(.) to have option to impute missing data, and changing that to be the default behavior

Fix bug when adding missing variances (PR 36)
Fix bug in detecting high ratio of variances (issue 35)
Added test_dsep(.) to calculate a p-value for the probability that a a data set arose from the specified model (highly experimental)
Added total_effect(.) to calculate total effects
Fixed bug where dsem crashed when quiet = FALSE and running without data e.g., as qualitative network model
Added vignette showing how to approximate diffusive movement using DSEM with paths among adjacent strata

Edits to allow DSEM to run when all path coefficients are fixed (contributed by Maurice Goodman)

Adding option for lower and upper bounds
Adding stepwise_selection for automated stepwise model selection
Adding plot option to use ggraph as alternative to previous igraph option
Adding convert_equations to extend sem::specifyEquations and simplify specification for large models
Adding argument prior_negloglike as interface to specify Bayesian priors and/or likelihood penalties in dsem
Adding dsemRTMB using RTMB as interchangeable alternative to dsem using TMB

Adding option to specify constant marginal variance, as alternative to existing calculation which results in a constant conditional variance but a changing marginal variance along the causal graph

removing checkDepPackageVersion(dep_pkg="Matrix", this_pkg="TMB") from .onLoad() as requested by K. Kristensen

Adding option to specify covariance via argument covs
Adding options to specify gmrf_parameterization=“projection”
Adding vigette outlining how to fit dynamic factor analysis
Fix bug arising when tsdata had two or more columns sharing a single variable name
Adding make_dfa helper function
Updating bering_sea dataset to include extra year of cold-pool, and changing vignette to match the published specification and results
Updating the lag indexing for the Klein-I model in the vignette to use positive values for lags, and updating saved MCMC results to match that corrected specification

Eliminate eval usage
Add codecov Action and badge
Change default behavior so that all variables in tsdata have a standard deviation by default