The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

iglm 1.2.6

Descriptive and Goodness-of-Fit Capabilities

• Unified Degree Distributions: The separate degree_distribution and spilloverdegree_distribution methods of class iglm.data have been merged into a unified, flexible degree_distribution(x_i, x_j, y_i, y_j, mode) method. Users can compute degree distributions on induced subnetworks where \(x_i, x_j, y_i,\) and \(y_j\) are constrained to specific values.
  • Attributes can be filtered using exact values (e.g., 0 or 1), discrete sets, continuous above-mean indicators (where \(x_i > \bar{x}\) for continuous/count covariates), or custom filtering functions.
  • Users can isolate directional spillover pathways: outgoing ties quantify spillover sending capacity, while incoming ties quantify spillover exposure.
• Extended Network Descriptive Distributions: The edgewise_shared_partner_distribution, dyadwise_shared_partner_distribution, and geodesic_distance_distribution methods of iglm.data now also accept x_i, x_j, y_i, y_j, and mode arguments to evaluate structural properties on attribute-conditioned subgraphs.
• Enhanced Goodness-of-Fit (assess): Because the assess method evaluates descriptive statistics specified on the left-hand side of formula terms, the entire goodness-of-fit toolbox seamlessly extends to these generalized, attribute-conditioned distributions.
• Global vs. Local Modes: Added the mode parameter ("global" by default, or "local") to all descriptive distributions, allowing users to restrict evaluations either to all network ties or solely to ties between units with overlapping neighborhoods (\(\mathcal{N}_i \cap \mathcal{N}_j \neq \emptyset\)).

Consistency and Alignment of Terminology

• family Parameter: Renamed the type parameter (type_x, type_y) to family (family_x, family_y) to align with standard R conventions (such as glm()) for specifying the conditional distributions of attributes.
• Unified Unit Terminology: Standardized terminology throughout the package code, function arguments (e.g., n_units replacing n_actor), and documentation to consistently refer to network entities as “units” rather than mixing “actors” and “units”.
• Shorthand Distribution Aliases: Added convenient shorthand methods to the iglm.data class:
  • deg_dist() for degree_distribution()
  • geo_dist() for geodesic_distance_distribution()
  • esp_dist() for edgewise_shared_partner_distribution()
  • dsp_dist() for dyadwise_shared_partner_distribution()

Labeling of Attributes and Terms

• Custom Attribute and Network Labels: Users can now supply custom labels via label_x, label_y, and label_z when constructing iglm.data objects.
• Enhanced Output Formatting: summary() and print() methods for both iglm fitted models and iglm.data objects display informative, domain-specific labels instead of generic canonical identifiers (e.g., attribute_republican and spillover_republican_turnout), with options to toggle between labeled and canonical names.

Robustness and Input Validation

• Formula Scope Verification: Added environment checks ensuring that all objects referenced in model formulas exist in the calling environment before estimation begins.
• Numerical Safeguards: Strengthened optimization and estimation routines to prevent silent propagation of NA or NaN values.
• Model Term Validation (check.IglmTerm): Introduced comprehensive validation helpers for model terms, checking mandatory arguments, expected data types (scalars, matrices), allowed categorical values, and positional-to-named argument mappings.
• Descriptive Formula Traps: Added informative error messages when formulas are incorrectly specified (e.g., passing raw vectors or lists instead of an iglm.data object on the LHS).

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.