# network

This document describes the "network" C++ class. This class describes the network object, that is used as an entry point for all other objects.

Abstract | can contain | contained in |
---|---|---|

no | compartment | nothing (root) |

## Methods¶

### checkSolvers¶

**Function Signature**

void checkSolvers(void)

**Description**

This method verifies that all components can integrate
using the requested solver order. What this method does
is to call the `checkSolvers`

in every compartment,
which in turn calls the `checkSolvers`

method in every
component contained in every compartment.

**Code**

Click here to view this method's code

### resolveTree¶

**Function Signature**

bool resolveTree(void)

**Description**

This method "resolves" a multi-compartment neuron model.
What this means is that it works out which compartment
is "upstream" (closer to the soma) or "downstream" (further
from soma) for every compartment in a multi-compartment model.
It does so using the `tree_idx`

property in every compartment,
setting it if need be.

It returns `true`

if there is a multi-compartment neuron model
somewhere in the network, and `false`

otherwise.

**Code**

Click here to view this method's code

### ¶

**Function Signature**

mexPrintf("[C++] resolveTree() called\n");

**Description**

This method "resolves" a multi-compartment neuron model.
What this means is that it works out which compartment
is "upstream" (closer to the soma) or "downstream" (further
from soma) for every compartment in a multi-compartment model.
It does so using the `tree_idx`

property in every compartment,
setting it if need be.

It returns `true`

if there is a multi-compartment neuron model
somewhere in the network, and `false`

otherwise.

bool network::resolveTree(void) { compartment * connected_comp = NULL; bool is_multi_comp = false; if (verbosity > 0) {

**Code**

Click here to view this method's code

### addCompartment¶

**Function Signature**

void addCompartment(compartment *comp_)

**Description**

This method adds a compartment to the network. It does the following things:
1. adds a pointer to the compartment being added to a vector called `comp`

2. Broadcasts certain global parameters like `temperature`

, `dt`

, etc to all compartments.
3. Updates `n_comp`

to that network knows how many compartments there are.

**Code**

Click here to view this method's code

### integrateMS¶

**Function Signature**

void integrateMS(double * I_ext_now)

**Description**

This method is used to integrate the network using a multi-step Runge Kutta solver. This method assumes that no compartment is being voltage clamped.

**Code**

Click here to view this method's code

### integrate¶

**Function Signature**

void integrate(double * I_ext_now)

**Description**

This method is used to integrate the network using the default single step solver. Typically, this means using the exponential- Euler method to integrate conductances and integrate the voltages and Calcium levels in compartments, though mechanisms can implement their own integration schemes. Multi-compartment models are integrated using the Crank-Nicholson scheme.

This method assumes that no compartment anywhere is being voltage clamped.

**Code**

Click here to view this method's code

### integrateClamp¶

**Function Signature**

void integrateClamp(double *V_clamp)

**Description**

This method is used to integrate the network using the default single step solver. Typically, this means using the exponential- Euler method to integrate conductances and integrate the voltages and Calcium levels in compartments, though mechanisms can implement their own integration schemes. Multi-compartment models are integrated using the Crank-Nicholson scheme.

This method assumes that some compartment is being voltage clamped, and also assumes that no current is being injected into any compartment.

**Code**

Click here to view this method's code