# New link

## Introduction

Links in a Bayesian network or Dynamic Bayesian network, indicate that the variable(s) in one node, influence the variable(s) in the other.

The distribution for a node, depends on the incoming links of a node. More specifically, the distribution for a node `A`

is the conditional probability `P(A|parents(A))`

.

Some properties are only shown if relevant, such as those that relate to DBN nodes, as shown below.

## Swap

The **Swap** button simply changes the **From** node(s) to the
**To** node(s), and vice versa.

## Existing distributions

When a link is added, the variables in a node's distribution change. If distributions have been specified before the link is added, the behavior depends on the option Migrate Existing Distributions.

## Temporal links

When creating a link between two temporal nodes in a Dynamic Bayesian network,
a **Temporal Order** must be specified. A temporal order indicates that the
**From** node links to the **To** node at a future time. A value of zero,
indicates that they are linked at the same time.

The **From** node and **To** node, can be the same node, if the node is a Temporal node, and the temporal order is greater than zero.

Multiple links can be created between two temporal nodes (including the same node) if the temporal order is different.

To add multiple links with different orders in a single operation, the **Range** option can be set to true,
allowing the input of a minimum and maximum order. Multiple links will then be added, from the minimum to the maximum order (inclusive).

## Continuous Nodes

While a link can be specified from a Discrete node to a Continuous node, the reverse is not true. i.e. a link cannot be specified from a Continuous node to a discrete node.