Deeprust Codebase
Machine learning crate in Rust
Installing the package
or alternatively you could install it by mentioning the library name in the Cargo.toml dependecies section
Basic Example
- Import the library
- Use the API's available according to the application that you are trying to build, for example I want to get the confusion matrix of my model's performance
#[macro_use(Confusionmatrix)]
extern crate deeprust;
use deeprust::metrics::confusion_matrix;
fn main() {
// assigning random values to the confusion matrix
let sample = Confusionmatrix!(100,50,10,5);
println!("The total predictions {}", sample.total());
// Calculating the accuracy of the model
println!("Accuracy of the model {:.2}", sample.accuracy());
// Calculating the precision of the model
println!("Precision of the model {:.2}", sample.precision());
// Calculating the true positive rate of the model
println!("True positive rate of the model {:.2}",
sample.true_poitive_rate());
// Calculating the false positive rate of the model
println!("False positive rate of the model {:.2}",
sample.false_positive_rate());
// Calculating the misclassification rate of the model
println!("Misclassification rate of the model {:.2}",
sample.misclassification_rate());
// Calculating the specificity of the model
println!("Specificity of the model {:.2}", sample.specificity());
// Calculating the prevalance of the model
println!("Prevalance of the model {:.2}", sample.prevalance());
}
For more details on the different APIs of the crate head over to the official documentation There are more examples over there.
Setting up the project locally
- Clone the repo
// Downloading the project
git clone https://github.com/dvigneshwer/deeprust
cd to_proj_dir
cd ./code/deeprust
- Run the build script
Additional info:
-
To get a mathematical gist of all the implementation here, check out this link
-
If you are new to Rust language, please feel free to checkout the rust training dir
-
Stuck at any place or spotted an issue with the project or do you have a feedback, feel free to to raise an issue
-
We are always looking for contributors,checkout the issue section & join the gang