No normalization in the inverse_haversine

[yoramzarai]

Hi

Thank you for your code. I find it very useful.

I do have two questions:

1. In the haversine function, you either normalize the latitude and longitude or check that they are in the correct range. However, in the inverse_haversine function, after computing the output latitude and longitude, you by default do not normalize them, and there is no checking that they are in the range. Why is that?
2. How would you use the haversine and the inverse_haversine functions to map, for example, two GPS points (each defined by its latitude and longitude) to 2 points on a Cartesian 2D (XY) grid, and then back to GPS points?

[jdeniau]

Hi,

In the haversine function, you either normalize the latitude and longitude or check that they are in the correct range. However, in the inverse_haversine function, after computing the output latitude and longitude, you by default do not normalize them, and there is no checking that they are in the range. Why is that?

By default, we do not normalize the values neither in "haversine", nor "inverse_haversine" (but it may be the default in the next major version)

You are right though, we do not check that initial lat/lng are correct in inverse_haversine, it might be added in another PR.

How would you use the haversine and the inverse_haversine functions to map, for example, two GPS points (each defined by its latitude and longitude) to 2 points on a Cartesian 2D (XY) grid, and then back to GPS points?

I do not know what you are trying to achieve, and I might not be the best help you can here, but if I remember correctly, projecting a sphere onto a map requires to make some choice in how you make your projection, isn't it?