NEWS

---

PartialNetwork 1.0.4 (2024-05-15)

Adjustments with CDatanet 2.2.0

PartialNetwork 1.0.3 (2024-03-06)

Adjustments with Eigen 3.4

PartialNetwork 1.0.2 (2023-08-22)

• We provide a smoother simulator of adjacency matrices in the SGMM approach.
• We add weights to the probit/logit network formation model.
• We allows the use of an initial probit/logit estimate of $\rho$, where the observed part of the network is assumed non-stochastic in the MCMC. This is a quite different from using an initial probit/logit estimate as prior distribution of $\rho$. In this latter case, $\rho$ is updated using, among others, information from the observed part of the network. In the first case, $\rho$ and the unobserved part of the network are updated using information in $y$, where the initial estimate acts as prior distribution of $\rho$. Information from the observed part of the network is not used to update $\rho$. This information is included in the initial estimate.

PartialNetwork 1.0.1

• As shown in the paper, we propose a simulated general method of moments (SGMM) for the SAR model (see function smmSAR and Section 2 of our vignette).
• We can now estimate the maximal bias of the instrumental variable estimator (see Section 1.1 and 1.2 of our vignette).

PartialNetwork 1.0.0

First release