Copyright (C) 2021 Parker Edwards
Command line R applications for machine learning using the R package TDAExplore.
Install TDAExplore-ML using the included script install.R. If installing from the command line R interpreter, make sure that this package's root directory is the working directory (using getwd() and setwd() in R). Then run:
R
> source("install.R") Alternatively, from RStudio, check that the working directory is this package's root directory. Then run in the following command in the console:
source("install.R")On Windows you will need to add R.exe and Rscript.exe to your PATH to run the batch scripts properly. See here for instructions.
On Linux several of the R package dependencies of TDAExplore-ML require that your system has certain C/C++ libraries installed. The links below are to appropriate Ubuntu packages to highlight the correct library. This list is not exhaustive: recursive package requirements may require more libraries. If library installations throw such an error, search through the logs for the right library, install it, then rerun the TDAExplore-ML install script.
The package includes two command line applications: ml-tda and convolve-tda. On Windows these are batch scripts ml-tda.bat and convolve-tda.bat. For both, pass the --help flag to see a listing of command line flags. The applications require a .csv parameters file with path indicated by --parameters. A template is provided that you can copy, rename, edit, etc.
ml-tda --parameters parameters_template.csv --svm TRUE --cores 6 If you do not want to relocate the scripts or put them on your PATH, you can always call them directly:
TDAExplore-ML/ml-convolve --parameters parameters_template.csv --svm TRUE --cores 6Images for the examples are distributed separately. You can download the data here and place the folder in this project's root directory to execute the examples.
TDAExplore is licensed under GPLv3.