Simple Fortran Math Problem
I am asked to follow the path of a column using GPS in near real time and compare the results to some other results.
The GPS gives you an ASCII file with Easting, Northing, time etc in a standard form. Fortran can read it easily and do the trivial math.
This is the raw data from a simple GPS and it has not yet been post processed, with RTKLIB written in C. I was just looking at the raw.
A sample output from 8 hours of eastings yields
It is trivial to put it into XY but there is no need here, the ground theoretically covered is a line about 5 metres long, as noted here
If I collect the data over 2 days I get a mess like this
which is a 2.5 metre circle of data.
But now I have 80000 points, about 1 per second and the data is following a set of patterns related to the satellites and the screening with a period of about 4 hours, the question is how stable is the center and how long to do you have to record to get a stable center estimate.
But if I break the 80000 points into 10000 point sets and do math, I have covered the cycle period nicely each time and the results suggest about 1 mm of total movement, about what I would expect. The graph is in mm.
Am I merely seeing a manifestation of the central limit theorem?
Why are the GPS programs not written in Fortran, it is a pain.
I will look with RTKLIB now to see how close that gets.
Link Copied
