Jump to content

co-ord averaging


Recommended Posts

If the degrees part of the co-ords are the same, and it's just the minutes parts that are different then stick just the minutes part into your favourite spreadsheet and average them out.

 

E.G N50 51.123, W001 01.456 & N 50 52.014, W001 00.983

 

just calculate (51.123+52.014)/2 and (01.456+00.983)/2

 

and when you have the average values put the N50 and W001 parts back.

Link to comment

I know how do do that but when you have a lot to do software would soon breeze through it.

 

That's why I suggested doing it in a spreadsheet. In excel if you have a list of co-ords in a text file, you could open it as a delimited text file, choosing space and comma as the delimiters, which will break up each part of the co-ords into different columns, then use the "average" function to do the maths, e.g. if you enter into a cell "=average(a1:a23)" (without the quote marks) it will work out the average of all the values in the first 23 cells of column A.

Link to comment

I don't have excel,only open office don't think it has the same options.

 

:)

 

I use OpenOffice too (at home), but just made the assumption about excel as that seems to be what everyone else uses. google says OpenOffice Calc has the Average function too, but I can't check it from work. Another way you could work out the average is to do "=sum(A1:A23)/23" and I know Calc does have the sum function, but that requires changing the formula based on how many rows you're averaging)

Link to comment

Oy....

 

OpenOffice

LibreOffice

Apple Numbers

Google Docs

 

They're all compatible with Microsoft Excel. Same formula syntax, same formulas. Sometimes there are multiple formula options for the same operation to be backward compatible with older versions of each of the other software packages.

 

The centroid of a cluster of points in n dimensions is the vector of the means of each of the n dimensions. In two dimensional space, the center of a cluster of points with coordinates (x_i, y_i) is (mean(x), mean(y)).

 

Personally, I do all of my calculations in R.

Edited by mineral2
Link to comment

The coordinate averaging subject reminds me of the previously discussed question of what is the basis of a GPSr's EPE value. All units display a value for each location; however, AFAIK, there has not been a provision of a unit's statistical definition of EPE value. For example, how may readings at that location would be within the displayed EPE? 50%? One standard deviation? Etc.?

 

OTOH, a user can determine that relationship by coordinate averaging.

Link to comment

The coordinate averaging subject reminds me of the previously discussed question of what is the basis of a GPSr's EPE value. All units display a value for each location; however, AFAIK, there has not been a provision of a unit's statistical definition of EPE value. For example, how may readings at that location would be within the displayed EPE? 50%? One standard deviation? Etc.?

The answer varies slightly between manufacturers and their specific proprietary formulas, but generally the displayed EPE ("Estimated Position Error") on a device means that 50% of position readings are statistically likely to fall within that range. Note that this also means that 50% of readings could be outside of that range.

 

If you double the EPE figure, it indicates that 95% of readings will fall within that doubled range.

 

If you want to spend a few hours reading some dry statistical discussion, there's lots of information available if you Google "estimated position error". :laughing:

Link to comment

The coordinate averaging subject reminds me of the previously discussed question of what is the basis of a GPSr's EPE value. All units display a value for each location; however, AFAIK, there has not been a provision of a unit's statistical definition of EPE value. For example, how may readings at that location would be within the displayed EPE? 50%? One standard deviation? Etc.?

The answer varies slightly between manufacturers and their specific proprietary formulas, but generally the displayed EPE ("Estimated Position Error") on a device means that 50% of position readings are statistically likely to fall within that range. Note that this also means that 50% of readings could be outside of that range.

 

If you double the EPE figure, it indicates that 95% of readings will fall within that doubled range.

 

If you want to spend a few hours reading some dry statistical discussion, there's lots of information available if you Google "estimated position error". :laughing:

 

I've tested by revisiting dropped waypoints, both with and without averaging.

 

of course averaging is closer, but without us plenty good enough, unless is some land formation that produces bounces.

 

honestly these things are pretty good.

Link to comment

Which simply says that if closer is better, averaging is preferred.

 

As for EPE (The-A-Team), we've seen some simply absurd EPE figures being provided by certain cell phone apps (1 foot or less), so I don't put much stock in any reading where the use of 'within x feet y % of the time' isn't provided as a hard spec.

The use of 50% is the normal 'CEP' description for EPE. At one time, it seemed that this was the measurement Garmin was displaying, but I haven't tried to pin it down on anything they've made recently. Personal experience says that it may currently be a bit better than 50%.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...