Bug in Train view/Workout Wizard?

[EricH]

Since I am doing Vineman 70.3 this year I decided to try to get the
bike course into GC so I could ride it in training. So I went to
BikeRouteToaster and mapped out the course, saved it as a GPX file,
and then loaded it as an activity into GC.

Possible bug #1: In bikeroutetoaster the course is 56 miles long but
when I loaded it into GC it was only 51 miles. Where did the other 5
miles go?

I loaded the activity into GC and then went into the Workout Wizard to
convert it to a crs file. The resultant file was then loaded into the
Train View. I know the UI is still a work in progress so I won't nit
pick on axis labels etc, but ...

Possible bug #2: When a crs course is first loaded it always starts
out on the left side as "200" (the axis is labeled Watts, but in slope
mode I would think it will eventually be elevation) and then, even if
the course is net uphill, the profile ends much lower than the start.
When you move the "elevation" slider at the top even just a little bit
the profile snaps to a more reasonable (accurate?) overall slope.

Possible bug #3: the slider on the top for this course has the
elevation at 226 (don't know what this refers to BTW) at the middle
but when I slide it all the way to the right it reads 378. If I click
on another crs file and then click back to the vineman.crs, the slider
at the middle reads 168 and at the far right reads 252.

All of these are repeatable and are consistent even with different gpx/
crs files loaded from BikeRouteTosater and Garmin Connect. I do not
think there is anything special about my Vineman course.

Thanks,
Eric

[Greg Lonnon]

Eric,

Can you do a bug report and add the gpx file.

Or, just send me the gpx file and I will dump your email and gpx file into bugs and look at it this weekend.

Thanks,

Greg

--
_______________________________________________
Golden-Cheetah-Users mailing list
golden-che...@googlegroups.com
http://groups.google.com/group/golden-cheetah-users?hl=en

[EricH]

I would like to add that when I rode the generated file I quickly
realized the distance was not 51 miles, but rather 51 km. Anyway, this
was Greg's response to my issue that I thought I would put here to see
if anyone has some good ideas:

Eric,

The GPX distance isn't encoded in the file, we calculate the distance
based on the GPS information in the file. We are using the haversine
formula to figure out the distance.

http://en.wikipedia.org/wiki/Haversine_formula

I have re-implemented the distance calculation using slightly
different constants from this website:

http://blog.julien.cayzac.name/2008/10/arc-and-distance-between-two-points-on.html

But, it didn't fix the 10% error.

I had a hunch that your file didn't have enough data to be accurate.
The file only had 816 track points, and the Haversine formula does a
point to point calculation. Thus, it doesn't follow the road. I took
a few of my tcx files that are sampled every second and converted them
to gpx files (using connect.garmin.com). The result was an error >
20%. It got worst!!!

Hmmm. I am out of ideas at this point. If there is a GPS expert and/
or math/trig whiz... Please help us out... The mapmyride and garmin
connect websites get the correct mileage from the gpx files.

Sorry,
Greg