The main purpose of this package is to perform sparse estimation of many normal means. Estimation is done using either the hard- or the soft-threhold estimator. An iterative approach is used to select a threshold.
Set up is easily done through the devtools
package.
Run the following R-script to install itses
:
library(devtools)
devtools::install_github("AmiAhm/itses")
Alternatively, download package material or clone the repository, and run the following R-script. Ensure to have the main folder "itses/" in the working directory.
library(devtools)
devtools::install("itses")
Running the main method:
library(itses)
# To get data to work with
y <- rnorm(10)
# To run at default settings (noise level estiamted + soft-threshold + numerical)
itses.result <- itses::itses(y)
# Print result
print(paste("Optimal threshold is:", itses.result$lambda))
Refer to the documentation and vignettes for in-depth explanation of parameters and alternatives. E.g. by:
?itses::itses
The package uses no external dependencies and was built on R version 4.0.3.
Fix rogue linebreak in utility causing faulty HT risk
This package was made as in part of the research project component of the MSc in Statistics of Imperial College London.
Thesis: Sparse Estimation of Many Normal Means
Implementation used in thesis can be found at: https://github.com/AmiAhm/SMNM-Implementation
By Amir Ahmed, supervised by Alastair Young (September 2021)