Bonsai

Zapit TCP/IP comms Bonsai demo

Before you start

1. The Extensions folder should be in the same directory as the Bonsai_Demo file.
2. You should have started Zapit with the "tcpServer" "enable" setting being "true" in the YML settings file.

Available commands

The Bonsai TCP client makes available the following commands that interface with the Zapit server.

• Stop_Optostim
• Stim_Conf_Loaded
• Get_State
• Get_Num_Conditions
• Send_Samples

Only sendSamples accepts input arguments. The remaining commands are run without input arguments.

Connection between a Zapit server and MATLAB client on the same machine

By default the IP address on the Zapit server and client is localhost, meaning that communications happen on the local machine and not over the network. Here we communicate between Zapit and a Bonsai client running on the local machine.

We will start Zapit in simulated mode and "calibrate" in MATLAB.

>> start_zapit('simulated',true)

% "calibrate" stereotaxic coords
>> hZP.applyUnityStereotaxicCalib

% Show that no client is currently connected
>> hZP.tcpServer.isClientConnected
No client is connected to the TCP server

You will now start the TCP/IP client on the local machine. To do this, open the Bonsai workflow. For this example, we will reuire:

• the Demo_Variables workflow
• the Zapit_Client_IO workflow
• the Extensions folder (in this same repository)

Start by importing all of these modules into your workflow. These can be accessed thorugh the search bar on the top left, then pressing Enter (or double clicking on them).

To establish a connection, you will have to send a message to the server via the Connect workflow. One really simple way of achieving this is by importing the Connect node, placing it inside a SelectMany and triggerring it with a KeyDown such a C (see example below). To check that everything is working properly, import and visualise the Server_Reply Subscribe_Subject (think of this simply as a variable). Upon trigerring the Connect node succesfully, the server should reply: (Connected, (1, 1))

Since Zapit has just started, there is no loaded stimulus configuration. So, if you import and trigger the Stim_Conf_Loaded? node (wth method of your choice), the Server_Reply should print: (2023-09-29 16:04:04.458000, (0, 255)). Similarly, the number of conditions is zero. So trigerring the Get_Num_Conditions? should yield: (2023-09-29 16:04:04.458000, (0, 255))

Now we will load an example stimulus set through MATLAB (or you can use the GUI to load one of your choosing).

pathToExample = fullfile(zapit.updater.getInstallPath,'examples','example_stimulus_config_files');
exampleFiles = dir(fullfile(pathToExample,'*.yml'));
hZP.loadStimConfig(fullfile(pathToExample,exampleFiles(1).name))

Let's verify that it loaded in MATLAB.

>> hZP.stimConfig

ans = 

  stimConfig with properties:

            configFileName: 'C:\zapit\examples\example_stimulus_config_files\uniAndBilateral_5_conditions.yml'
            laserPowerInMW: 5
      stimModulationFreqHz: 40
             stimLocations: [1×5 struct]
    offRampDownDuration_ms: 250
               chanSamples: [25000×4×5 double]

Now triggering the Stim_Conf_Loaded? and Get_Num_Conditions? should yield (2023-09-29 16:04:04.458000, (1, 255)) and (2023-09-29 16:04:04.458000, (1, 255)) respectively.

We can then initiate stimulus delivery using Send_Samples on the client. By clicking on the node, we can set the parameters:

hardwareTriggered_channel = False
hardwareTriggered_ON = False
condition_Num_channel = True
conditionNum = 2

In which case the server should return (2023-09-29 16:04:04.458000, (2, 1)). We can stop stimulation by triggering the Stop_Optostim node, which should return (2023-09-29 16:04:04.458000, (1, 255)).

When finished you disconnect the client using the Disconnect node. Triggering this should yield (Disconnected, (0, 0)).