MC_Rack / MC_Stimulus II on Linux

Hi!

This is just to let other users know that I was successful in making subj. to run on Linux. You need Wine 1.2 package. Install the software as normal. It will not run. Download winetricks and select vcrun2008, that is latest Microsoft Visual C++ runtime. Follow the installation. Run the software again. Now it works.

I didn't test it thoroughly but being able to open your stimuli and replay your files is already something. It's quite annoying that the GUIs provided are Windows-only.

Best,
Z.

FYI, I was equally successful in running DataTool ASCII conversion. It turned out, however, that the output format is not quite the same as described in the documentation, and if you are recording spike cutouts, you need to post-process the resulting file to extract spike times. I can post the script if anyone is interested.

Is the output format for spikes different on Linux compared to Windows?

And where is the difference as the spike format seems not to be described in the documentation? Only raw data tables are shown, spikes could not be exported in this format, as each channel has different numbers of entries and timestamps. We will include this information into the documentation.

Jens

Hi!

Jens wrote:Is the output format for spikes different on Linux compared to Windows?

No, it's the same, otherwise this would be a Wine-related bug.

Jens wrote:And where is the difference as the spike format seems not to be described in the documentation? Only raw data tables are shown, spikes could not be exported in this format, as each channel has different numbers of entries and timestamps. We will include this information into the documentation.

See "Converting Data to ASCII Format", ASCII file output. This section is about exporting raw data, which I assumed was also applicable to spike cutouts. Instead of being arranged in columns as a matrix, the channels are written to the file sequentially and the cutouts seem to be separated by a newline. This is not specifically mentioned anywhere and had to be discovered by trial and error.

I don't see how the fact that the number of items in each column might be different matters as long as they are TAB or CSV-like (comma or semicolon) separated. Empty values are detected by stumbling upon consecutive semicolons. Check how OpenOffice Spreadsheet exports CVS data with empty cells.

Best,
Z.

We will discuss, if we could consider you proposal in one of the next versions of MC_DataTool and how we could implement it in a compatible way for users expecting the current format.

Jens

Jens wrote:We will discuss, if we could consider you proposal in one of the next versions of MC_DataTool and how we could implement it in a compatible way for users expecting the current format.

Well, the biggest problem is that it is not documented. I already wrote a parser for your current format to obtain the matrices of spike timestamps vs. electrodes out of this, so I'm fine now as long as it doesn't change :-)