# Conceptual problem with user recipes: draw a sphere before drawing the data
**URL:** https://discourse.julialang.org/t/conceptual-problem-with-user-recipes-draw-a-sphere-before-drawing-the-data/49385
**Category:** General Usage
**Tags:** plots, recipe
**Created:** [October 31, 2020, 8:12pm UTC](https://discourse.julialang.org/t/conceptual-problem-with-user-recipes-draw-a-sphere-before-drawing-the-data/49385 "2020-10-31T20:12:53Z")
**Posts on this page:** 6
**Page:** 1
### Author:  [@kellertuer](https://discourse.julialang.org/u/kellertuer)
#### Post date: [October 31, 2020, 8:12pm UTC](https://discourse.julialang.org/t/conceptual-problem-with-user-recipes-draw-a-sphere-before-drawing-the-data/49385/1 "2020-10-31T20:12:53Z")
I am trying to do a user recipe from [RecipesBase.jl](http://juliaplots.org/RecipesBase.jl/stable/) but I seem to not get the idea how Recipes work I fear.
**Context**.
I would like to do something that I thought was quite easy. Given the `M=Sphere(2)` from [Manifolds.jl](https://github.com/JuliaManifolds/Manifolds.jl) and a set of points (maybe just `pts=[[ 1.0, 0.0, 0.0], [0.0, 1.0, 0.0] ]` I would like to be able to write
`scene = plot(M,pts)`
to draw a sphere and two markers where these points are,
Best would be that for the scene I could also add things using `plot!(scene, M, pts2)` for a second set of points.
Here’s my current approach – and I can’t find what I am doing wrong, I fear I have a conceptual problem with recipes still.
Here’s my current (nonworking) MWE
```julia
using Manifolds, Plots, RecipesBase
@recipe function f(
M::Sphere{2},
pts::AbstractVector{T};
wireframe_lat=33,
wireframe_lon=33,
) where {T}
show_axis --> false
φ_range = range(0,stop=2*π,length=wireframe_lon)
λ_range = range(0,stop=π,length=wireframe_lat)
x = [cos(φ) * sin(λ) for φ ∈ φ_range, λ ∈ λ_range]
y = [sin(φ) * sin(λ) for φ ∈ φ_range, λ ∈ λ_range]
z = [cos(λ) for φ ∈ φ_range, λ ∈ λ_range]
scene = plot() # I assume this should be the global plot? Wherefrom?
wireframe!(scene, x, y, z)
seriestype --> :scatter
return [p[1] for p ∈ pts], [p[2] for p ∈ pts], [p[3] for p ∈ pts]
end
M = Sphere(2)
pts = [[1.0, 0.0, 0.0], [0.0, 1.0, 0.0] ]
plot(M, pts)
```
and I also was not able to find something similar on the net.
How do I add something to a plot, when usually the recipe only returns `x, y, z` and I can’t seem to access the plot to add the sphere? It just plots the two points.
---
### Author:  [@ggarza](https://discourse.julialang.org/u/ggarza)
#### Post date: [October 31, 2020, 8:55pm UTC](https://discourse.julialang.org/t/conceptual-problem-with-user-recipes-draw-a-sphere-before-drawing-the-data/49385/2 "2020-10-31T20:55:39Z")
After a very quick glance at your code, I think you have a problem in the line with the following:
```julia
y = [sin(lon) * sin(lat) for φ ∈ φ_range, λ ∈ λ_range]
```
`lon` and `lat` are not defined.
---
### Author:  [@kellertuer](https://discourse.julialang.org/u/kellertuer)
#### Post date: [October 31, 2020, 9:05pm UTC](https://discourse.julialang.org/t/conceptual-problem-with-user-recipes-draw-a-sphere-before-drawing-the-data/49385/3 "2020-10-31T21:05:26Z")
Thanks for noticing, sorry, I copied an old version – name them `φ` and `λ` and it works, the plot actually has both points in, just no sphere
I also edited the original post.
---
### Author:  [@kellertuer](https://discourse.julialang.org/u/kellertuer)
#### Post date: [November 2, 2020, 9:10am UTC](https://discourse.julialang.org/t/conceptual-problem-with-user-recipes-draw-a-sphere-before-drawing-the-data/49385/4 "2020-11-02T09:10:16Z")
I found one trick that seems to be a step forward – ` @series` from
but I can’t do a surface with that as I would do with `surface(x,y,z)` below
using Manifolds, Plots, RecipesBase
```julia
@recipe function f(
M::Sphere{2},
pts::AbstractVector{T};
wireframe_lat=33,
wireframe_lon=33,
) where {T}
show_axis --> false
φ_range = range(0,stop=2*π,length=wireframe_lon)
λ_range = range(0,stop=π,length=wireframe_lat)
x = [cos(φ) * sin(λ) for φ ∈ φ_range, λ ∈ λ_range]
y = [sin(φ) * sin(λ) for φ ∈ φ_range, λ ∈ λ_range]
z = [cos(λ) for φ ∈ φ_range, λ ∈ λ_range]
@series begin
# seriestype := :surface # this line does not work if you provide a surfce as set of points?
x,y,z #plots a line per z value not a surface
end
seriestype --> :scatter
return [p[1] for p ∈ pts], [p[2] for p ∈ pts], [p[3] for p ∈ pts]
end
M = Sphere(2)
pts = [[1.0, 0.0, 0.0], [0.0, 1.0, 0.0] ]
plot(M, pts)
```
what I would like the `@series` is something either similar to `wireframe` or `surface` (in the end I want to be able to activate one, the other or both).
_Edit_: Ah, I see, so what I would like to obtain is what Make(!) does when you call `surface(x,y,z). on those data generated above, I.e. a nice surface of the sphere. How can I do that in Plots? And don’t recipes also work for Make? But they are still so different?
---
### Author:  [@kellertuer](https://discourse.julialang.org/u/kellertuer)
#### Post date: [November 8, 2020, 4:03pm UTC](https://discourse.julialang.org/t/conceptual-problem-with-user-recipes-draw-a-sphere-before-drawing-the-data/49385/5 "2020-11-08T16:03:17Z")
So my main problem seems to be that the series can not be convinced to do a 3D sphere (half transparent as a surface or as a mesh, either would be fine – later I want to be able to activate one of those or both). The result could vor example be similar to the plots I do with exports and rendering (Asymptote) [here](https://manoptjl.org/stable/tutorials/MeanAndMedian.html).
So how can I get a surface sphere in a series here ? Neither `:surface` nor `:wireframe` as `series type`do work.
---
### Author:  [@kellertuer](https://discourse.julialang.org/u/kellertuer)
#### Post date: [November 9, 2020, 3:00pm UTC](https://discourse.julialang.org/t/conceptual-problem-with-user-recipes-draw-a-sphere-before-drawing-the-data/49385/6 "2020-11-09T15:00:18Z")
I found the culprit: Ith works with `plotly()` as backend but not with `gr()`.
_edit:_ I opened an Issue [here at Plots.jl](https://github.com/JuliaPlots/Plots.jl/issues/3126).