Averaging

[phern47]

I was out ice fishing the other day and had good luck fishing through my ice fishing hole. I had cut a hole through the ice in exactly the right spot. When I let my 2 ounce lead lure and bait, I could feel the lure bounce along the side of a large boulder. The area where I was fishing is very large and quite flat. There are hundreds of ice shantys and fishermen all over the place but I wanted to return exactly to the same fishing hole because the fish seemed to congregate near this large boulder.

 

I powered up the IFinder Pro that I had just bought "used" and tried out the option of "averaging" before creating my waypoint. I waited about 15 seconds and had 30 points registered and then I created my waypoint. When I went out the next day, I quickly drove my vehicule right next to my original hole even though the wind and snow had almost obliterated all evidence of a fishing hole.

 

Do any other GPS units have this option? I seem to remember my Garmin 12 having such a similar function. Should I have waited even longer before creating my waypoint? How accurate does this make the GPS? Any opinions about this function?

Edited January 21, 2008 by phern47

[+Chuy!]

You used the function as it was intended.

 

My Vista HCx and Venture Cx have it. But I create the waypoint first, then use the averaging option to fine-tune the coordinates.

[+JohnTee]

My original Garmin eTrex Legend does not have that capability. With it I had to mark multiple coordinates and then use Thot's Geocaching Combo to average my multiple coord's.

 

As mentioned by Chuy! my new Garmin Vista HCx does a nice job of averaging coordinates. I've not tried creating a waypoing then fine-tuning it however. I'll hit the 'Mark Location' button, then go right down to average for 20-100 cycles, then hit save.

 

Cache On!

 

JohnTee

[yogazoo]

Garmin Colorado - Does not!

[+W7WT]

When I set GZs for the GPS Accuracy Game we run a couple time a year, I used my 76C with a external amplilfied antenna at 15 feet and averaged for a long time. What was interesting was that the indicated accuracy went down to about 2.5 feet but the coordinates didn't change. I just wanted to be sure I got it as accurate as possible. I also only did it when I had a good satellite pattern which I believe is very important. dick

[+fjones15]

Yes my DeLorme PN-20 dose. and I used it for the same thing.

[+Team DeLorme]

Yes my DeLorme PN-20 dose. and I used it for the same thing.

 

Make sure you get the 1.4 firmware update; we just improved this feature so that you can return to a location you have averaged and gather new data on a different day. You will see the number of readings continue to increase and the accuracy estimate update. This is valuable since the GPS constellations change with time and your best averages are gathered over multiple visits.

 

For those using our XMap GIS software you will also find averaging tools within the GPS tab in the software.

 

Let me know if you have any questions...

 

Chip Noble

Team DeLorme

[Hertzog]

Make sure you get the 1.4 firmware update; we just improved this feature so that you can return to a location you have averaged and gather new data on a different day. You will see the number of readings continue to increase and the accuracy estimate update. This is valuable since the GPS constellations change with time and your best averages are gathered over multiple visits.

You should also suggest that they return at different times of the day, since the constellations at a given location pretty much repeat every 24 hours.

 

Regarding the achieved improvement in accuracy with averaging: I don't know what your algorithms for the PN-20 are, but I'm pretty sure the Garmins reduce the accuracy by something like 1/sqrt(number of samples) during averaging; this would be correct if the samples were uncorrelated, but that isn't the case for samples taken over a short period of time. (They may not be doing quite that, but my feeling is that their reduced error estimates are too optimistic.)

 

It's good to see a manufacturer come online and interact with customers this way. While I am still a committed Garmin fan (their customer support is outstanding) they tend to standoffish and noncommittal about the workings of their products. The next time I upgrade I will definitely be looking at the DeLorme products, along with the Garmins.

[+apersson850]

It's even better to use a prediction software, to compute when the satellite constellation is best. Leica Geosystems have one, for example, that can be downloaded for free. There are others, too.

 

Thread

Leica

[+fjones15]

[Make sure you get the 1.4 firmware update;

Chip

When I downloaded it there were only 6 hits on the announcement. Much faster routing. Getting 8 blue bars inside the house through is slate roof. Hope to do some side by side with other units this week.

fred

[Hertzog]

...I'm pretty sure the Garmins reduce the accuracy by something like 1/sqrt(number of samples) during averaging...

After reviewing some tests I made about a year ago, I have to retract my statement about Garmin's use of a simplistic 1/sqrt(#samples) reduction in estimated error. I'm not sure precisely what they are doing, but the results are not as dramatic as you would get with a 1/sqrt reduction.

[+Team CowboyPapa]

Having a capability and having a need for that are two different things. For example, I've seen people drive their Lexus 4WDs to the end of the pavement and then turn around and go back to town.

 

Having set the scene with that comment, I'll be more pertinent. I have the PN-20 and use its averaging function when placing a cache.

 

OTOH, I don't use that capability when searching for a cache. Am I missing something there?

Edited January 22, 2008 by CowboyPapa

[jmundinger]

But I create the waypoint first, then use the averaging option to fine-tune the coordinates.

 

I'm missing something here. Just how do you go about "fine tuning" a waypoint after it is created? What is the sequence?

 

My basic approach, when hiding a cache, is to create one averaged waypoint at the cache site, walk a few feet away from the hiding spot, return and create a second averaged waypoint. Time permitting, I might do a third. Then, I plot all three in USAPhotoMaps and pick the waypoint that looks to be the best.

[+apersson850]

I don't remember now where I saw that you can more or less proove mathematically that averaging waypoints, when you don't know anything about the circumstances, doesn't make you any better than you were.

[+Team DeLorme]

But I create the waypoint first, then use the averaging option to fine-tune the coordinates.

 

I'm missing something here. Just how do you go about "fine tuning" a waypoint after it is created? What is the sequence?

The key is in returning to the same location over several days, at different times of day, to update your GPS average. Returning over several days increases the random nature of your readings... reduces the influence of the constellation pattern and the atmospheric interference. Taking one reading or multiple readings within the same visit means that the error associated with the constellation and atmosphere are the same for all readings.

 

The Earthmate GPS PN-20 keeps track of the total number of readings as well as the change in accuracy as new data is collected. After creating the initial waypoint just return to the location, view the Waypoint Details page and press Menu to select the Average Waypoint... option. This will continue averaging the location you previously recorded. Keep in mind that this process is only as good as the data collection pattern used.

 

We recommend putting your GPS device to the test... use a benchmark with a known location. Create a waypoint for your known location, download it to your desktop application, draw a circle around it with radius equal to the error... if the waypoint error covers the known location then the device is reporting an accurate error value. Repeat this test over multiple days, at different times of day, and you will see the intersection of these circles begin to define a smaller area that represents the average of your readings. I’ve done this test using the PN-20 and high resolution aerial imagery from our area and have been pleased with the results.

 

Here is a link to some very informative GPS accuracy and averaging research by David Wilson, a mathematician from Kent State University...

 

http://users.erols.com/dlwilson/gps.htm

 

Let me know if you have any questions about getting the best location information for your geocaches...

 

Chip Noble

Team DeLorme

[jmundinger]

The Earthmate GPS PN-20 keeps track of the total number of readings as well as the change in accuracy as new data is collected. After creating the initial waypoint just return to the location, view the Waypoint Details page and press Menu to select the Average Waypoint... option. This will continue averaging the location you previously recorded. Keep in mind that this process is only as good as the data collection pattern used.

 

Thanks for the explanation. I haven't tried it, but I don't think the Vista HCx has the option of doing what you just described, i.e. averaging the waypoint after the waypoint has been created - I think the average function only works when the waypoint is first created and before it is saved, and, thus, my confusion. It sounds as though the PN-20 has that capability.

 

Oooops. I just learned something new about my GPSr - it does have the capability of averaging a waypoint after the waypoint has been created - it's a function on the menu for the waypoint.

Edited January 23, 2008 by jmundinger

[Smac999]

just playing with my HCX. learning new stuff . never used or looked for the avg before.

 

say I avg a waypoint. and let it count to 100. and save it.

 

if I go back to that point, as mentioned above I can find the avg waypoint option again. but it starts couting at 1 again. I wonder if it's simply trying to recalc the pos. or if it's actully using any data from the first try (which it sounds like the PN-20 does)?

[Smac999]

well I just tryed something. I avged a waypoint with 20 counts with my HCX. I moved 40 feet away, and re avged it with 20 more counts.

 

I simply ended up with a point 40 feet away from the first. had it avged the way I would have hoped I should have ended up with the new point 20 feet in the middle?

 

so I don't think the HCX can be used to combine multiple data on different days. it'll simply just move the point to where it finds it that day. and won't care were it was before.

 

that is my take from my quick 30 sec test anyways.

Edited January 23, 2008 by Smac999

[Hertzog]

well I just tryed something. I avged a waypoint with 20 counts with my HCX. I moved 40 feet away, and re avged it with 20 more counts.

 

I simply ended up with a point 40 feet away from the first. had it avged the way I would have hoped I should have ended up with the new point 20 feet in the middle?

 

so I don't think the HCX can be used to combine multiple data on different days. it'll simply just move the point to where it finds it that day. and won't care were it was before.

 

that is my take from my quick 30 sec test anyways.

As you observed, the Garmins restart the averaging every time you average. But this is a better approach anyway; just mark and average a new waypoint each time you want to make a measurement, and average them offline; the timestamps in the waypoints give you a record of when you made your measurements, and you can get a feel for how much variation you are getting from measurement to measurement.

[+Ratsneve]

Yes my DeLorme PN-20 dose. and I used it for the same thing.

 

Make sure you get the 1.4 firmware update; we just improved this feature so that you can return to a location you have averaged and gather new data on a different day. You will see the number of readings continue to increase and the accuracy estimate update. This is valuable since the GPS constellations change with time and your best averages are gathered over multiple visits.

 

For those using our XMap GIS software you will also find averaging tools within the GPS tab in the software.

 

Let me know if you have any questions...

 

Chip Noble

Team DeLorme

If you average and can drive right up to your fishing hole / cache one day why do you have to keep averaging the location any more? Over time you mention that the "GPS constellations change". What is this and why does it matter if you already have a locations coordinates?

Edited August 21, 2008 by Ratsneve

[yogazoo]

I don't remember now where I saw that you can more or less proove mathematically that averaging waypoints, when you don't know anything about the circumstances, doesn't make you any better than you were.

 

Right, provided your GPS doesn't jump position at the exact wrong time. In the open, with a great view of the sky I could agree HOWEVER if your in a canyon/heavy tree cover and your position is jumping around at the order of a few dozen meters then things might be different. That test you speak of, what kind of environmental conditions were present? I'll bet that they didn't even document it or only used a single type. In this case that test would be misleading or only giving you half of the story. Any time there is sampling you have statistics and if your not careful with your variables you could be distorting the truth.

 

On a related note, Delorme seems to be listening to serious users who use their GPS's as a tool rather than a toy. Delorme PN-40 is going to give Garmin a run for their money with serious navigational and professional users. At least they interact with users in this forum. It means alot not to be held at arms length as Garmin does to it's customers who have problems.

 

And to answer the question of the thread: EVERY OUTDOOR HANDHELD GPS DEVICE HAS WAYPOINT AVERAGING CAPABILITIES EXCEPT THE GARMIN COLORADO AND OREGON LINES!!!

Edited August 21, 2008 by yogazoo

[+JSWilson64]

And to answer the question of the thread: EVERY OUTDOOR HANDHELD GPS DEVICE HAS WAYPOINT AVERAGING CAPABILITIES EXCEPT THE GARMIN COLORADO AND OREGON LINES!!!

 

Really? Come on down to Dallas, and I'll pay for your plane ticket if you can show me how to do it on my Etrex Legend. And if you can't do it, you can buy me a plane ticket to Hawaii.

 

For those with handhelds that don't average, but can connect to a computer, you can average waypoints yourself. Take several waypoints (the more the better) at the site you want to average. Return at different times, if possible and repeat. Then download the waypoints to Easy GPS or your favorite program. Convert coords to decimal degrees, import into Excel (or your favorite spreadsheet). Average the Latitude. Average the Longitude. Done.

 

My son and I used an Etrex Legend in heavy tree cover to scout sites for a possible cache near our home. We did the above process, collecting 15-20 waypoints over 3 trips to the site. The averaged coords, entered into a 60CSx, take us right to the tree we were considering using. Just haven't gotten around to placing the cache...

[yogazoo]

And to answer the question of the thread: EVERY OUTDOOR HANDHELD GPS DEVICE HAS WAYPOINT AVERAGING CAPABILITIES EXCEPT THE GARMIN COLORADO AND OREGON LINES!!!

 

Really? Come on down to Dallas, and I'll pay for your plane ticket if you can show me how to do it on my Etrex Legend. And if you can't do it, you can buy me a plane ticket to Hawaii.

 

 

Sorry Tex, you're right, the Legend can't average waypoints. Let me rephrase the statement by saying that every handheld GPS introduced for serious outdoor use within the past 5 years has waypoint averaging. The Legend was summer of 2002. That includes all of the Garmin Color mapping units, Magellan Meridian lines, explorist lines, Triton Lines, Lowrance I-finders, Delorme models, Trimble models, etc. You'd be hard pressed to find a real serious navigation / data collection device today that doesn't include Waypoint Averaging. I guess that was my point.

 

Your solution about taking a mess of waypoints and manually averaging them sounds enticing but I'd sooner have my unit do the averaging. I knew I paid 500 big ones for a Colorado for some reason, and it wasn't to play cheezy geo-games with Wherigo.

 

As for your offer? I really do like Texas. Spent 3 years at Ft. Hood and loved it. I was there during G-Dub's term as governor and I still liked it. BUT, I'll stay in Montana, thanks.

Edited August 21, 2008 by yogazoo

[+Team DeLorme]

If you average and can drive right up to your fishing hole / cache one day why do you have to keep averaging the location any more? Over time you mention that the "GPS constellations change". What is this and why does it matter if you already have a locations coordinates?

You only need to average if you're concerned that your coordinate accuracy is low. If you're able to drive up to your fishing hole or cache I'd say your finished. People who are averaging require a high degree of accuracy for some reason... placing a cache is a good example.

 

By changing GPS constellations I just meant that the satellites you can see on any given day change as do the atmospheric conditions. These are variables that contribute to the strength of your GPS signal and the accuracy of your readings.

 

If you want a high level of confidence in your geocache location I would recommend averaging the cache location until you see that your GPS Accuracy has stabilized at a low value... then I'd repeat the averaging process on multiple days, at different times of day.

 

You should test this in your back yard. Average GPS readings for the corner of your driveway or some other repeatable location. Do this on multiple days... When you are finished I believe you will see several waypoints that are close but not exactly at the same location. Averaging these points will give you a better estimate of the measured coordinates.

 

I hope this clarifies things for you.

 

Chip Noble

Team DeLorme

[+StanByk]

 

Chip Noble

Team DeLorme

 

Hi Chip from Team DeLorme,

 

I am considering to get a PN-40 once it comes out, however, I am based in Europe (London, UK). Will the PN-40 be available on the European market. Will I have access to topographical as well as (routable) streetmaps for the UK as well as the rest of Europe?

 

Thank you

Robert

[+Team DeLorme]

 

Chip Noble

Team DeLorme

 

Hi Chip from Team DeLorme,

 

I am considering to get a PN-40 once it comes out, however, I am based in Europe (London, UK). Will the PN-40 be available on the European market. Will I have access to topographical as well as (routable) streetmaps for the UK as well as the rest of Europe?

 

Thank you

Robert

 

Unfortunately the only product that is available for Europe is our XMap software and it does not offer street level detail for road routing. This will allow you to load "raster" maps... scanned paper maps or aerial imagery like you see on Google... Our pro-sales department has more information on this, as well as a world dataset, but it will not give you access to the level of street information that I think you're asking for. Requests like yours certainly help us prioritize future projects, thank you for asking.

 

Chip Noble

Team DeLorme

[+StanByk]

 

Chip Noble

Team DeLorme

 

Hi Chip from Team DeLorme,

 

I am considering to get a PN-40 once it comes out, however, I am based in Europe (London, UK). Will the PN-40 be available on the European market. Will I have access to topographical as well as (routable) streetmaps for the UK as well as the rest of Europe?

 

Thank you

Robert

 

Unfortunately the only product that is available for Europe is our XMap software and it does not offer street level detail for road routing. This will allow you to load "raster" maps... scanned paper maps or aerial imagery like you see on Google... Our pro-sales department has more information on this, as well as a world dataset, but it will not give you access to the level of street information that I think you're asking for. Requests like yours certainly help us prioritize future projects, thank you for asking.

 

Chip Noble

Team DeLorme

 

Many thanks for your reply. What I can hear from these boards here, Delorme seems to have a very good reputation. Should Delorme ever expand to the European market I would seriously consider to buy a Delorme unit.