Calculate radius and remove GPS coordinates - gps

Question

I have a question about how to remove some "unnecessary" coordinates from a "maps.txt" file using segment line, projection vector or other method.
Path Created on Google Maps:
Map1
Total of 275 coordinates, extracted from the ".KML" file of Google Maps.
Map1_2
Map2
When I do it manually it stays that way (Map2), with 9 coordinates.
I have a file called "maps.txt" where its lines, with coordinates, are in this format:
Obs: A coordinate on each line: latitude, longitude.
-37.2012600, -59.8404600
-37.2000200, -59.8419600
-37.1985300, -59.8439200
-37.1970600, -59.8458500
-37.1959100, -59.8473500
-37.1957800, -59.8475200
-37.1948600, -59.8486900
-37.1939500, -59.8498600
-37.1931400, -59.8509400
-37.1928400, -59.8513100
-37.1926700, -59.8515000
-37.1924600, -59.8517200
-37.1922600, -59.8519200
Is there any way for a code to read my file "maps.txt" and do the line segment, vector projection, point distance calculation using a radius of 100 meters?
In case the code would "remove the line / unnecessary coordinate" from the file "maps.txt", leaving it this way (just an example):
-37.2012600, -59.8404600
-37.1948600, -59.8486900
-37.1922600, -59.8519200
In Python, C, C ++ or another language.
I hope I have been clear and thank you in advance for any help.
Thank you (:
I'll try to explain better.
In the following figure I have 3 coordinates (1, 2 and 3).
The coordinate "1" has a radius within 100m with respect to the "2" coordinate, but from "1" to "3" I have a value greater than 100m, ie in this case the coordinate "1", "2" "and" 3 "in my text file" maps.txt ".
Example1
In the following figure from example 2, I would need only the coordinates "1" and "3" in the file "maps.txt".
Example2

Answer 1

Update:
Here's a script that converts KML to GPX: https://gist.github.com/timabell/8791116
And here's a Python script that seems like it'll do what you want (for GPX files): https://wiki.openstreetmap.org/wiki/User:Travelling_salesman/gpx_reduce
Previous answer:
I don't know how far you're asking us to take our answers; this problem is slightly complex, so here's some pseudocode for starters:
Overall algorithm:
Vector from point 1 to 3. Check distance of point 2 from this vector. (Assuming threshold is okay)
Vector from point 1 to 4. Check distance of points 2 and 3 from this vector. (Assuming threshold is okay)
Vector from point 1 to 5. Check distance of points 2, 3, and 4 from this vector. (Assuming threshold is bad)
First vector is determined: it's from point 1 to 4. Second vector starts at point 4.
Vector from point 4 to 6. Check distance of point 5 from this vector.
Vector from point 4 to 7. Check distance of points 5 and 6 from this vector.
...
Distance of point from vector:
At this point, we need to make a decision:
If we want this to be (quite) exact, we have to find a line that is perpendicular to our line and goes through this point, and find the intersection of these two lines. Then, we calculate the distance using the haversine formula between the intersection and the point. (Unless we sometimes have longer distances between coordinates, I don't think we need this.)
If we don't care too much about exactness, we use the following formula to calculate the "distance", and then "experimentally" find a good threshold value that works for us. (This threshold value is just some float number that we use when checking the distance. If it's above the threshold value, we start a new line.)
https://en.wikipedia.org/wiki/Distance_from_a_point_to_a_line#Line_defined_by_two_points
(Where (x0, y0) is the point we're investigating, and the line goes through (x1, y1) and (x2, y2).)