why does Diadem always have issues with csv files?

[RichElswick]

Yet again, I am having an issue with Diadem 10.0 opening a .csv file.  Yes,  the format may not be the perfect .csv file, but the old adage goes...

Excel opens it.  If NI wants to get Excel users to migrate, this is one of those items, where an Excel user would be annoyed and not even make the effort to change.  I tried writing a Plugin (not simple) and it did not work out for me.

I have attached an example .csv file, which I am having issues with.

Thanks for the help.

[Brad_Turpin]

Hello Rich,

As best I can tell, the default CSV file reader that ships with DIAdem is stumbling over the simple linefeeds at the end of your file's lines (CRLF is more standard).  Everything else looks just like it should expect it, though you would get an empty channel at the end.  I'll forward that on to R&D for fixing.

In the meantime, try this DataPlugin I created for you.  Let me know if this somehow also gives you trouble.

Cheers,
Brad Turpin
DIAdem Product Support Engineer
National Instruments

[RichElswick]

this did it for me, thanks!

[RichElswick]

Well, I thought this had done it for me and it did pull in the file with the column headers being mapped to the channel headers, but the first column being time did not pull in the values correctly.  Such as the time would show 100, 120 180, 2, 220, 260, 320, 4, 450

where 2 and 4 should have been 200 and 400.  It could be the larger file sizes, but I am not sure.  I will post 2 files in a zip file named Rich_Cobasys.zip to the ftp site for you to take a look at.  The time series (first channel in each group) when imported should match up.

Let me know if you are getting the same errors.

Thanks,
Rich

[RichElswick]

By the way Brad,

Could you delete the file once you download it?

Thanks.

[Josh_W]

Rich,

The FTP server is cleaned every few days, and only internal NI employees are able to download information from the /incoming folder.

With the three data files we have now, and the data plugin posted above, we have been unable to reproduce the incorrect time values you have been seeing.  Are you seeing this behavior with any of these three files, or with other files? 

Since we have been unable to reproduce the behavior, we don't know exactly what's going on.  So if you could send along a screenshot of the behavior that you are seeing, we will be able to assist you better.  If you are seeing this behavior with a file other than the three we have, if you could also send along that file, it would help a lot.

Josh W. | National Instruments | Applications Engineering

[RichElswick]

cool.  Here is the screen shot of the last .zip files uploaded.  Note, you have to divide the non _CAN file channel 1 by 100 via 'ch(1):=ch(1)/100)' if you import that file first to get this screenshot.

this is weird.

[Josh_W]

Rich,

We've got a few ideas about what might be going on, but since we can't reproduce the error here, we're really just guessing.

Something that might cause trouble is that the default data loading settings might be different between our machines and your machine.  So the new attached data plugin has a few more settings explicitly defined.  The one we're most concerned about is the property to ignore spaces within numeric cells.

I didn't see anything wrong with the Total Time channel you included in your screenshot.  What should I be looking for?

Josh W. | National Instruments | Applications Engineering

[RichElswick]

I will try out the new plugin tomorrow.

Take a look at the attached screenshot with circles to highlight the details.  260 seconds should match up with 260, but it shows 2 instead.  At the bottom, the Total Time column shows 365 and this is what should be in the same row for Total Test Time, but it shows 300,000 instead.