https://github.com/sciqlop/speasy.jl
A simple Julia wrapper around Speasy
https://github.com/sciqlop/speasy.jl
Last synced: about 1 month ago
JSON representation
A simple Julia wrapper around Speasy
- Host: GitHub
- URL: https://github.com/sciqlop/speasy.jl
- Owner: SciQLop
- License: mit
- Created: 2025-01-26T10:08:14.000Z (4 months ago)
- Default Branch: main
- Last Pushed: 2025-04-04T21:18:29.000Z (about 2 months ago)
- Last Synced: 2025-04-04T21:31:13.939Z (about 2 months ago)
- Language: Julia
- Homepage:
- Size: 50.8 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Speasy
[](https://github.com/Beforerr/Speasy.jl/actions/workflows/CI.yml?query=branch%3Amain)
A simple Julia wrapper around [Speasy](https://github.com/SciQLop/speasy), a Python package to deal with main Space Physics WebServices.
## Features
- Integration with [`DimensionalData`](https://github.com/rafaqz/DimensionalData.jl) (recommended), [`TimeseriesToolsExt`](https://github.com/brendanjohnharris/TimeseriesTools.jl), [`TimeSeries`](https://github.com/JuliaStats/TimeSeries.jl), and [`Makie`](https://github.com/MakieOrg/Makie.jl).
## Installation
```julia
using Pkg
Pkg.add("Speasy.jl")
```## Examples
```julia
using Speasy
const spz = speasyget_data("amda/imf", "2016-6-2", "2016-6-5")
# Dynamic inventory
amda_tree = spz.inventories.data_tree.amda
get_data(amda_tree.Parameters.ACE.MFI.ace_imf_all.imf, "2016-6-2", "2016-6-5")
``````julia
# More complex requests
products = [
spz.inventories.tree.amda.Parameters.Wind.SWE.wnd_swe_kp.wnd_swe_vth,
spz.inventories.tree.amda.Parameters.Wind.SWE.wnd_swe_kp.wnd_swe_pdyn,
spz.inventories.tree.amda.Parameters.Wind.SWE.wnd_swe_kp.wnd_swe_n,
spz.inventories.tree.cda.Wind.WIND.MFI.WI_H2_MFI.BGSE,
spz.inventories.tree.ssc.Trajectories.wind,
]
intervals = [["2010-01-02", "2010-01-02T10"], ["2009-08-02", "2009-08-02T10"]]
get_data(products, intervals)
```> [!NOTE]
> It is advisable to load this package before any others, as it relies on OpenSSL underpinnings. Compatibility issues may arise between Python and Julia if it is not prioritized accordingly.