Page 5 of 8

Posted: Fri May 29, 2015 1:39 pm
by bmotorsports
Looking at the logs, distance starts at some crazy number like 5778 for me. Then it starts a lap by lap count of distance, resetting each lap.

Anyway, you can go to Data at the top in GEMS, then distance channel setup. Unclick the first box and go to the second box and you can calculate distance from speed. You have to add the speed channel. The distance variable has a problem in the RCP logs and every time I tried to use the GPS data, GEMS just crashes.

Posted: Fri May 29, 2015 2:07 pm
by JMcDonough
My experience echo's bmotorsports's. I have been post processing (in excel) the distance data from the GPS speed. At 50Hz, the raw distance data coming out of RCP was not accurate. I switched to 10Hz, which has been giving better position data, and creating my own distance data channel based on speed in Excel after converting the file with the perl script from this thread. I then use Dlog99 to convert and configure the distance channel after opening in GEMS.

All my data is from auto-x, so I can't speak to how lap by lap distance works.

In summary, I'm longing for the day ASL (re)releases their analysis software. Hopefully it will have good functionality like GEMs.

Posted: Fri May 29, 2015 6:11 pm
by brentp
Thanks for the update.

Next time (or during a test run on the streets) can you try the built-in distance algorithm using 10Hz as a reference?

We're working hard on the in-app analysis. Here's a screenshot from a while back - it's rough but will give you an idea of usage.



Image[/img]

Posted: Sat May 30, 2015 9:34 am
by mavab91
Hi All,

I'm trying to use the tool RCP2GEMS with the example file but something is not working.
Steps I did.

1. git clone https://github.com/autosportlabs/RCP2GEMS.git
2. cd RCP2GEMS/
3. ./python/RCP_conv.py RCP_Demo.LOG RCP_Demo_dlog.csv
its output was:
0 lines skipped
(True, 'RCP_Demo_dlog.csv')

Then I opened Dlog99 (I downloaded it from http://www.gems.co.uk/?content=pages&id ... -downloads) I chose File Import and selected RCP_Demo_dlog.csv, Float Import, but I am getting the "0.0 is not a valid floating point value" (I attached a picture of the error message).

Do you have any idea of what could be wrong?

Thanks in advance.

Posted: Sat May 30, 2015 2:09 pm
by JMcDonough
Hi mavab91, I want to say I've seen that error before as well.

2 suggestions to try:

1. If that file is open in another program (Excel), close that program and try the import again.

2. Try opening the .csv file in Excel, then re-saving (maybe with a new name) as a csv and closing.

Hope this helps.

Posted: Sat May 30, 2015 3:20 pm
by mavab91
I have managed to solve this issue, I had my computer regional settings set to Spanish (Colombia), I changed it to USA and everything is working.
(Spanish (Colombia) uses ,commas as decimal separator).

Posted: Mon Jun 01, 2015 5:27 am
by mat.walker17
Thanks Bm and Jc,

That would also explain the strange gps speeds I was getting, in the range of 80 km/h slow at an actual 200km/h.

Do we know of anyone in Australia using this hardware?

Posted: Tue Jun 02, 2015 12:59 am
by parid
First off, big thank you to everyone who has contributed to the GEMS conversion script. It was very straight forward to use.

I got two stints from our last ChumpCar race imported all the way and we have made some great progress already learning things from the data. There are a couple things really driving me nuts. I can't get segments working.

1. I can't get the segments to work at all. I have the distance channel setup and appears to be working well.
2. I have the track generated via GPS.
3. I plotted a start beacon.
4. I have laps now split up and lap times calculated.

I have tried both auto generated segments and manually inputted segments via distance. No matter what, it always detects all of my data in just the first segment.

Any ideas what I'm doing wrong? I worry that its something messed up in the data that no amount of fiddling with the interface can fix.

Posted: Sat Jun 27, 2015 10:35 pm
by bigbcraig
All,

Hopefully simple question / answer, but I can't get the script to work... I'm by no means a programmer so I'm sure it's user error.

I've installed Perl from: http://padre.perlide.org/download.html , and as described in the RCP3GEMS Readme installed Text::CSV, and am now recieving the following error while trying to run the script on either my own logs or the example posted on github.

Does anyone know what's going wrong here?

Thanks,
Ben

Posted: Sun Jun 28, 2015 4:05 pm
by toga94m
Try opening your RCP_demo.LOG file in notepad - I bet it's really an HTML file. Did you rightclick a link and say SAVE AS or just click the link? Try getting the real log file, and it should be better. Or I could be completely wrong. :)

Posted: Mon Jun 29, 2015 11:22 pm
by bigbcraig
Great thinking, and thank you very much!

I actually had the .pl script saved as an html, fixing that and the script seems to have worked!

Nicer graphs in GEMS?

Posted: Sun Sep 06, 2015 12:02 pm
by fmbo
Does anyone know, is it somehow possible to tell GEMS DA that channel is sampled with different rate?

I don't see a way to change the default setting, and the chart created based on 1KHz sample rate for the channel which was actually logged at 10 Hz looks not very nice: http://prntscr.com/8d7h4z

Posted: Tue Oct 13, 2015 6:06 pm
by jojojones
I mentioned RC on here https://youtu.be/SmsqfNCmuYo

Posted: Thu Oct 22, 2015 8:29 am
by eangelov
I am trying to use the perl script to convert a *.csv file I acquired from my datalogger (my datalogger is not a RCP one) and it seems like it can't recognize the time column in the *.csv. The first column in my file is the time column. I attach some photos with the *.csv and cmd with error.

Posted: Fri Oct 23, 2015 11:33 pm
by JMcDonough
eangelov wrote:I am trying to use the perl script to convert a *.csv file I acquired from my datalogger (my datalogger is not a RCP one) and it seems like it can't recognize the time column in the *.csv. The first column in my file is the time column. I attach some photos with the *.csv and cmd with error.
Since the perl script was written specifically to convert RCP's .log files to .csv's that can be imported into DLOG99, you aren't going to have any success running the perl script on your data that is already in .csv form. You might try importing your .csv into dlog99 directly and see if that works.