Skip to content

maierhofert/CoFD

Repository files navigation

CoFD

Classification of Functional Data

Downloading the repository

To clone the repository, use:

git clone --recurse-submodules https://github.com/maierhofert/CoFD  # or git://

Overview of repository

  • benchmark_repo: Code for the benchmark. This is a git submodule. This repository contains all the code and descriptions for reproducibility. See benchmark_repo/README.md.

  • Benchmark_results: Folder containing the Benchmark Result (.bmr) files created by the benchmark experiments‌. The state with the latest date stamp is used to produce figures in code.

  • Data: Folder containing the data from the UCR TSC repository

  • Literatur: Miscellaneous papers that are related to the paper. Not a complete list of all cited papers

  • Plots: all plots generated and downloaded from the internet for the paper

    • BeetleFly: Original images of beetles and flies, extracted outlines for one beetle and one fly
    • EnsemblePartialDependence: partial dependence plots for the RFE and the LCE for all base learner combinations
    • benchmark: plots for results of benchmark experiments
  • R: R code

    • old: Unused R code. Can be deleted before publicising
    • Section3, Section4: R code chunks that are in paper
    • UCR_TSC_data: read in UCR TSC data, transform it to mlr tasks, select feasible data sets
    • benchmark_paper_server: Run this file on the server to execute the benchmark experiment
    • benchmark_paper: Code for setting up benchmark
    • benchmark_ressults_paper: Plot benchmark results, see /Plots/benchmark/
    • create_learners_paper: Create base learners, ensembles, and reference learners for benchmark
    • main_semimetrics: Overview of most important semimetrics currently implemented in classiFunc, their hyper parameters, and their possible range.
    • plot_Ensemble: Create LCE and RFE for subsequent plotting
    • plot_EnsemblePartDep_TE, plot_EnsemblePartDep_TE, plot_EnsemblePartDep_TE: create partial dependence plots for how the ensembles (tree, LCE, RFE) depend on their base models, see /Plots/EnsemblePartialDependence
    • plot_beetlefly: Create plots for transition of image to outline to 1d function of BeetleFly data
    • plot_classiFunc_users: Plot number of downloads of mlr and classiFunc per day
    • plot_kernelFunctions: Plot all available kernel functions
    • plot_model_weights: Create bar plots for the LCE (model weights) and RFE (variable importance)
  • CoFD.Rproj: R project. Open this and the working directory and all paths are correctly set for all R files in /R

About

Classification of Functional Data

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published