https://github.com/kmarkert/empiricalorthogonalfunctions.jl

Julia package for calculating Empirical Orthogonal Functions from spatiotemporal datasets.
https://github.com/kmarkert/empiricalorthogonalfunctions.jl

julia-language signal-processing spatial-temporal-data

Last synced: 3 months ago
JSON representation

Julia package for calculating Empirical Orthogonal Functions from spatiotemporal datasets.

• Host: GitHub
• URL: https://github.com/kmarkert/empiricalorthogonalfunctions.jl
• Owner: KMarkert
• License: mit
• Created: 2022-01-10T02:30:37.000Z (about 3 years ago)
• Default Branch: main
• Last Pushed: 2022-01-14T22:57:04.000Z (almost 3 years ago)
• Last Synced: 2024-10-11T14:41:38.749Z (3 months ago)
• Topics: julia-language, signal-processing, spatial-temporal-data
• Language: Julia
• Homepage: https://kmarkert.github.io/EmpiricalOrthogonalFunctions.jl/dev/
• Size: 4.67 MB
• Stars: 6
• Watchers: 2
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# EmpiricalOrthogonalFunctions.jl

Julia package for calculating Empirical Orthogonal Functions from spatiotemporal datasets.

This package was heavily inspired by the [`eofs` Python package](https://github.com/ajdawson/eofs) and a good amount of code was translated to Julia from this package.

## Installation

```julia

using Pkg

Pkg.add("EmpiricalOrthogonalFunctions")

```

##  Example

This example will highlight extracting the spatial and temporal flooding signals from a series of satellite imagery over Southeast Asia

```julia

using EmpiricalOrthogonalFunctions

using NCDatasets

#load in the data

ds = NCDataset("sar_stack.nc","r");

datain = ds["VV"][:];

# apply EOF

eof = EmpiricalOrthogonalFunction(datain; timedim=3)

# rotate the EOFs using varimax rotations

nmodes = 4

reof = orthorotation(eof,n=nmodes)

# extract out the signals

# the spatial signals are reshaped back to the original dimensions

temporalsignal = pcs(reof)

spatialsignal = reshape(eofs(reof),(size(datain)[1:2]..., nmodes))

```

When plotting the first four spatial signals we will get the following plot.

![](docs/src/assets/spatialmodes.png)

Below is the temporal signals corresponding to the spatial signals above

![](docs/src/assets/temporalmodes.png)