glide ratio

[trsonn]

Hello everyone,

I was playing a bit with python and some flysight data the last days, with the goal to input a flysight file into the software and to get out a distance in a certain altitude window.

Basically i want to check how my result would have been here.

Now i found some helpful resources online:
1. geopy and distance.py
2. the haversine formula
and 3. flajt

Now i did some thinking and tested various ideas to calculate the glide ratio in the competition altitudes.

With the haversine and vincenty formulas and with an average glide ratio of all the points in the competition video i am still just close (in the best case) to the result i get from paralog.

Do you guys maybe have an idea how paralog is doing these calculations?



Regards,

Tim

[trsonn]

sorry forgot:

http://competitionws.appspot.com/upload

The dropzone from my recordings is 135m above msl and this is hardcoded until now

[Michael Cooper]

Hi Tim—

 

I’m not sure what algorithm Paralog uses. However, there are some difficulties in calculating this kind of distance which you may not have thought of…

 

Perhaps the most obvious thing would be to calculate distance along the path between the start and end of the competition window. This would give you your “total” distance traveled. However, I think there is a major issue inherent in this calculation, described by the “coastline paradox”:

 

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

 

Essentially, there is no well-defined notion of “length” for the path between two points. Unfortunately, the total distance measured in this way depends sensitively on the where you choose to measure the individual positions. Our flight paths aren’t truly fractal, but they can have features which are small relative to the distance between consecutive points, so my feeling is that the coastline paradox applies. I suppose, as an experiment, that one could try calculating distance using various “granularities” to see if the distance converges to a single value.

 

Assuming there is no well-defined notion of path length, the best solution, I think, is to calculate the distance between the point where you entered the competition window and the point where you left it. This is the “straight-line” distance, and it is shorter than any “total” distance you might choose. This makes it the most conservative measurement of distance, which I think is a nice quality for competitions and records. Calculating glide ratio from the straight-line distance and the size of the competition window would again give you the most conservative measurement.

 

Hope this helps!

 

Michael

--
You received this message because you are subscribed to the Google Groups "FlySight Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flysight-dev...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Tim Sonnenberg]

Coastline Paradox...Yeah, now if feel kind of stupid to not have
thought of this. I think if i find a cool way to get rid of points
that are "off" i know what to do...

Thanks a lot!

Tim

2013/12/8, Michael Cooper <mic...@flysight.ca>:
> Hi Tim-

>
>
>
> I'm not sure what algorithm Paralog uses. However, there are some
> difficulties in calculating this kind of distance which you may not have

> thought of.

> <http://wingsuitboogie.hu/en/competition> .

>
> Now i found some helpful resources online:
> 1. geopy and distance.py
> 2. the haversine formula
> and 3. flajt
>
> Now i did some thinking and tested various ideas to calculate the glide
> ratio in the competition altitudes.
>
> With the haversine and vincenty formulas and with an average glide ratio of
> all the points in the competition video i am still just close (in the best
> case) to the result i get from paralog.
>
> Do you guys maybe have an idea how paralog is doing these calculations?
>
>
>
> Regards,
>
> Tim
>
> --
> You received this message because you are subscribed to the Google Groups
> "FlySight Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to flysight-dev...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
> --

> You received this message because you are subscribed to a topic in the

> Google Groups "FlySight Developers" group.

> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/flysight-devs/XNTnfHRxKbE/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to

> flysight-dev...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>

--
-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: OpenPGP.js v.1.20121007
Comment: http://openpgpjs.org
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=UNlq
-----END PGP PUBLIC KEY BLOCK-----