# Intersection of Ball and GeometrySet in LatLon crs

**URL:** <https://discourse.julialang.org/t/intersection-of-ball-and-geometryset-in-latlon-crs/123992>\
**Category:** Geo\
**Tags:** question, geometry\
**Created:** [December 19, 2024, 12:25pm UTC](https://discourse.julialang.org/t/intersection-of-ball-and-geometryset-in-latlon-crs/123992 "2024-12-19T12:25:19Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![stepanoslejsek](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/stepanoslejsek/32/218476_2.png) [@stepanoslejsek](https://discourse.julialang.org/u/stepanoslejsek)\
**Post date:** [December 19, 2024, 12:25pm UTC](https://discourse.julialang.org/t/intersection-of-ball-and-geometryset-in-latlon-crs/123992/1 "2024-12-19T12:25:19Z")

</div>

Hi, I’d like to find an intersection of a Ball with GeometrySet in LatLon coordinate reference system. However, when I construct my Ball `b` like

`b = GeoStats.Ball(GeoStats.Point(LatLon(48.34, 18.45)), 10)`

and ropes `r` as

`38 GeometrySet ├─ Rope((lat: 49.7838°, lon: 18.2359°), ..., (lat: 49.7869°, lon: 18.2258°)) ├─ Rope((lat: 49.7869°, lon: 18.2258°), ..., (lat: 49.7946°, lon: 18.2327°)) ├─ Rope((lat: 49.7946°, lon: 18.2327°), ..., (lat: 49.795°, lon: 18.2337°)) ├─ Rope((lat: 49.795°, lon: 18.2337°), ..., (lat: 49.7973°, lon: 18.2488°)) ├─ Rope((lat: 49.7973°, lon: 18.2488°), ..., (lat: 49.7974°, lon: 18.2495°)) ⋮ ├─ Rope((lat: 49.8325°, lon: 18.1762°), ..., (lat: 49.8329°, lon: 18.176°)) ├─ Rope((lat: 49.8329°, lon: 18.176°), ..., (lat: 49.8351°, lon: 18.1754°)) ├─ Rope((lat: 49.8351°, lon: 18.1754°), ..., (lat: 49.8353°, lon: 18.1747°)) ├─ Rope((lat: 49.8353°, lon: 18.1747°), ..., (lat: 49.8345°, lon: 18.1697°)) └─ Rope((lat: 49.8345°, lon: 18.1697°), ..., (lat: 49.8341°, lon: 18.17°))`,

then `b ∩ r` results in following error:  
`ERROR: StackOverflowError: Stacktrace: [1] intersection(f::Function, g₁::Ball{🌐, GeodeticLatLon{…}, Quantity{…}}, g₂::GeometrySet{🌐, GeodeticLatLon{…}, Geometry{…}}) @ Meshes C:\Users\oslejsek\.julia\packages\Meshes\K84Ms\src\intersections.jl:100--- the last 1 lines are repeated 79982 more times --- Some type information was truncated. Use `show(err)` to see complete types.`

Different types of projection didn’t help and resulted in the same error. Then I tried to make 2 balls close to each other and apply the intersection function, which results in same error.

Is there any workaround around this?  
PS: I am using GeoStats.jl v0.72.0 for this (but I guess the problem will be within Meshes.jl package).

---

<div class="post-metadata">

**Author:** ![juliohm](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/juliohm/32/215266_2.png) [@juliohm](https://discourse.julialang.org/u/juliohm)\
**Post date:** [December 19, 2024, 1:05pm UTC](https://discourse.julialang.org/t/intersection-of-ball-and-geometryset-in-latlon-crs/123992/2 "2024-12-19T13:05:49Z")

</div>

Hi @stepanoslejsek , thank you for sharing the issue. Geodesic geometries usually require custom geometric algorithms, which we plan to implement in the first semester of 2025.

It would be great if you could reduce the example to a pair with a `Ball` and a single `Rope`. Your example is trying to compute intersection between a `Ball` and a `GeometrySet` of `Rope`s.

I’d first try to check if the intersection works with `Cartesian` coordinates. Then, try again with `LatLon` coordinates. If the error message shows up again, please feel free to open an issue on GitHub.

---

<div class="post-metadata">

**Author:** ![stepanoslejsek](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/stepanoslejsek/32/218476_2.png) [@stepanoslejsek](https://discourse.julialang.org/u/stepanoslejsek)\
**Post date:** [December 19, 2024, 1:29pm UTC](https://discourse.julialang.org/t/intersection-of-ball-and-geometryset-in-latlon-crs/123992/3 "2024-12-19T13:29:01Z")

</div>

Sure, here is MWE for that:

```julia
using GeoStats

lats = [49.7868267, 49.7868428, 49.7868608, 49.7868831, 49.7869015, 49.7869172, 49.7869369]
lons = [18.2258339, 18.2258248, 18.2258177, 18.2258101, 18.225806, 18.225805, 18.2258069]

r = Rope([GeoStats.Point(LatLon(lat, lon)) for (lat,lon) in zip(lats,lons)])
# make center of the ball at one point of the rope
b = GeoStats.Ball(GeoStats.Point(LatLon(lats[5], lons[5])), 3)

intersection = b ∩ r

```

I tried `Cartesian` coordinates as well but the error is still persistent. I’ll open a github issue.

---

<div class="post-metadata">

**Author:** ![juliohm](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/juliohm/32/215266_2.png) [@juliohm](https://discourse.julialang.org/u/juliohm)\
**Post date:** [December 19, 2024, 1:32pm UTC](https://discourse.julialang.org/t/intersection-of-ball-and-geometryset-in-latlon-crs/123992/4 "2024-12-19T13:32:28Z")

</div>

Now I see the issue. You are trying to intersect a 2D ball wity a 1D rope. We don’t have a method for this pair yet, even for Cartesian coordinates and flat geometries.

---

<div class="post-metadata">

**Author:** ![juliohm](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/juliohm/32/215266_2.png) [@juliohm](https://discourse.julialang.org/u/juliohm)\
**Post date:** [December 19, 2024, 3:38pm UTC](https://discourse.julialang.org/t/intersection-of-ball-and-geometryset-in-latlon-crs/123992/5 "2024-12-19T15:38:20Z")

</div>

As a side note, notice that you don’t need to prepend all geometries with the `GeoStats` prefix unless you are loading two modules that export the same names. We usually import Makie backends with an explicit prefix instead, assuming that most of the geometric work is consistently handled by our stack:

```julia
import GLMakie as Mke

```

That way you can switch between Makie backends in a single place:

```julia
import CairoMakie as Mke

```

Regarding the issue above, we could start by adding an intersection method between `Segment` and `Ball`. The case with `Rope` could then be added with iteration over its `segments`.

The case with `Segment` and `Ball` can be as simple as checking a distance between the `center` of the `Ball` and the two `vertices` of the `Segment`. If any vertex is inside the `Ball`, you can then return a sub-`Segment`.

Please let us know if you plan to work on it. We can help with the PR review.

---

<div class="post-metadata">

**Author:** ![stepanoslejsek](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/stepanoslejsek/32/218476_2.png) [@stepanoslejsek](https://discourse.julialang.org/u/stepanoslejsek)\
**Post date:** [December 20, 2024, 8:18am UTC](https://discourse.julialang.org/t/intersection-of-ball-and-geometryset-in-latlon-crs/123992/6 "2024-12-20T08:18:29Z")

</div>

Not sure if I’ll manage to work on that, but I’ll try to manage it. Should I still open up an issue?

---

<div class="post-metadata">

**Author:** ![juliohm](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/juliohm/32/215266_2.png) [@juliohm](https://discourse.julialang.org/u/juliohm)\
**Post date:** [December 20, 2024, 8:37am UTC](https://discourse.julialang.org/t/intersection-of-ball-and-geometryset-in-latlon-crs/123992/7 "2024-12-20T08:37:26Z")

</div>

Please open an issue in the Meshes.jl repository. We will look into it.