Saqrlab is now simulation-only: network estimation code (build_network, bootstrap_network, permutation_test, temporal_network, mlvar, GIMME, MCML, higher-order networks, etc.) has moved to the sibling package Nestimate. This release focuses the package on synthetic-data generation with known ground truth.
saqr_sim interfacesaqr_sim S3 class wraps every numerical simulator's output with a
consistent $data / $params / $type / $seed interface, plus [,
as.data.frame(), head(), tail(), dim(), names(), str(),
print(), and summary() methods. $data / $params access is
backward-compatible.simulate() — single dispatcher that routes a type string to the matching
explicit-parameter simulator.list_simulators() — catalogs the available simulation types as a data.frame.validate_recovery() — compares a method's point estimates to the known true
params, returning a tidy per-parameter recovery table (absolute/relative
error and a within-tolerance flag).simulate_ttest(), simulate_anova(), simulate_correlation(),
simulate_clusters(), simulate_prediction() — generate data from
user-specified means, SDs, correlation/covariance matrices, cluster centers,
and regression coefficients; ground-truth effect sizes returned in params.simulate_longitudinal() — VAR(1) multilevel panel data for mlVAR/ESM with
explicit temporal, contemporaneous, and between-person matrices.inject_missingness() — adds MCAR, MAR, or MNAR missingness to any simulated
data frame at a controlled rate.simulate_mlm() and simulate_growth() — multilevel and latent-growth /
longitudinal data from explicit fixed effects and variance components.simulate_irt() — item response theory data (1PL/2PL/3PL) from explicit item
parameters and latent abilities.simulate_survival() — time-to-event data with explicit hazard, covariate
effects, and censoring.simulate_hmm() — hidden Markov model sequences from explicit emission and
transition probabilities.list_scenarios(), get_scenario(), and run_scenario() — named simulation
recipes for common designs.tidy_simulation_results() — flattens simulation results to a data.frame.export_simulation() — writes results to disk.simulate_group_tna_networks() (group TNA network generation).compare_networks() / network-comparison summaries.Initial release of Saqrlab - Simulation and Analysis Tools for Temporal Network Analysis.
simulate_matrix() - Generate simple transition matrices with learning state namessimulate_htna(), simulate_mlna(), simulate_mtna() - Multi-type matrices for hierarchical/multilevel network analysissimulate_sequences() - Generate Markov chain sequences with optional learning statessimulate_sequences_advanced() - Sequences with stability modes for realistic patternssimulate_long_data() - Hierarchical group data with actors, groups, and coursessimulate_onehot_data() - One-hot encoded format for sequence datasimulate_edge_list() - Social network edge lists simulationsimulate_tna_datasets() - Complete TNA datasets with sequences and parameterssimulate_tna_networks() - Fitted TNA models with random parameterssimulate_group_tna_networks() - Group TNA models for multi-group analysissimulate_tna_matrix() - HTNA/MLNA matrices with node typesgenerate_probabilities() - Random transition matrices and initial probabilitiesfit_network_model() - Fit TNA, fTNA, cTNA, and aTNA modelscompare_networks() - Compare two networks using correlation, RMSE, MAE, and morecompare_centralities() - Compare centrality profiles between networkscompare_edge_recovery() - Edge recovery metrics (precision, recall, F1)batch_fit_models() - Fit models to multiple datasets in parallelbatch_apply() - Apply any function to a list of objectsrun_bootstrap_simulation() - Bootstrap analysis for stability testingrun_grid_simulation() - Parameter grid search simulationsrun_network_simulation() - Comprehensive model comparison studiesrun_bootstrap_iteration() - Evaluate single bootstrap runssummarize_grid_results() - Analyze grid simulation outputget_learning_states() - Get learning verbs by categorylist_learning_categories() - Show available categories with countsselect_states() - Intelligent state selection based on network sizeLEARNING_STATES - Full dataset of 180+ learning verbs in 8 categoriesGLOBAL_NAMES - 300 diverse names for simulationget_global_names() - Retrieve names from the global names datasetgenerate_param_grid() - Create parameter combination gridsvalidate_sim_params() - Validate and set defaults for simulation parameterssummarize_simulation() - Summary statistics for simulation resultssummarize_networks() - Network-level summaries for model listsuse_learning_states = TRUE as default for realistic educational simulationsfuture and future.applyprogressrtna, seqHMMdplyr, tidyrparallel, future, future.applyprogressr, lhs