sanworks / bpod Goto Github PK

it seems that the ExampleMatrices should be in docs and not in the Functions folder. Because it is in the functions folder those matrices end up on the path. Seems like it is unintended.

Using a trinomial fit can result in very bad interpolation.

I recommend using linear interpolation and warning about volumes that are out of the linear range.

Hi,
I want to know how to do the water calibration, Can you explain the method? since the readme file is too short for me. or you can provide the reference paper.

Thank you

Using the bpod GUI users can create 'virtual' pokes. But I cannot emulate staying in a poke. The poke always immediately toggles back to un-poked. I would suggest that those buttons should act as toggle buttons (so that poking in and then poking out requires two clicks).

Yes, I yielded to MS insistence and accepted the upgrade to Win10, and found out Bpod-code no longer finds Bpod-serial device.
Specifically, the following happens in function FindArduinoPorts.m:

[Status RawString] = system('wmic, path Win32_SerialPort Where "Caption LIKE ''%Arduino%''" Get DeviceID')
RawString = 'No Instance(s) Available.'

Everything runs fine if I pass the port name as an argument to Bpod (e.g. Bpod('COM3')).

The CAD folder is only required for building boxes/PCBs not running Bpod. I would recommend moving that to another repo to save disk and download time.

In line with the issues raised by Jeff, I'd additionally suggest that Protocols be moved to a submodule, allowing it to be independently version-controlled.

Please post any feedback in our support forums. We like to keep the discussion in one place, so everyone can see it. Thank you kindly!!

What is the advantage of using run(ProtocolPath) vs. addpath(Protocolparentfolder); eval(ProtocolName)?

I understand that run effectively switches to the ProtocolPath, which gives the side-effect of having other files in that folder on the path, but that seems a bit weak, since if other calls switch directories those other files will no longer be on the path.

Seems safer to me to addpath before running the protocol and then rmpath at the end to avoid name conflicts.