https://github.com/dokato/natml

https://github.com/dokato/natml

Last synced: about 1 month ago
JSON representation

• Host: GitHub
• URL: https://github.com/dokato/natml
• Owner: dokato
• License: mit
• Created: 2022-04-05T12:12:51.000Z (almost 3 years ago)
• Default Branch: main
• Last Pushed: 2022-09-03T14:15:34.000Z (over 2 years ago)
• Last Synced: 2024-10-16T12:39:41.202Z (3 months ago)
• Language: R
• Size: 3.78 MB
• Stars: 1
• Watchers: 2
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# natML

Machine learning on neurons made easy. This package serves as an interface between the `nat` package for geometric neurons analysis and other R packages for machine learning. It let's you extract the features from your neurons seamlessly, regardless whether you want to access predefined columns, or calculate new features. It comes with extra functions returning some of the most popular scalar features for neurons classification.

## Installation

```r

# install.packages("remotes")

remotes::install_github("dokato/natML")

```

## Example

```r

library(nat)

library(natML)

# named or unnamed list of features

features_list <- list("upstream", "downstream", "voxels", "cellBodyFiber", get_cable_length)

kc_features <- extract_features(

  kc_train,

  features_list = features_list,

  to_numeric = TRUE,

  normalise = "zscore"

)

clusters <- kmeans(kc_features, 3)

```