Last updated: 2021-11-02

Checks: 7 0

Knit directory: 2019-feature-selection/

This reproducible R Markdown analysis was created with workflowr (version 1.6.2). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.


Great! Since the R Markdown file has been committed to the Git repository, you know the exact version of the code that produced these results.

Great job! The global environment was empty. Objects defined in the global environment can affect the analysis in your R Markdown file in unknown ways. For reproduciblity it’s best to always run the code in an empty environment.

The command set.seed(20190522) was run prior to running the code in the R Markdown file. Setting a seed ensures that any results that rely on randomness, e.g. subsampling or permutations, are reproducible.

Great job! Recording the operating system, R version, and package versions is critical for reproducibility.

Nice! There were no cached chunks for this analysis, so you can be confident that you successfully produced the results during this run.

Great job! Using relative paths to the files within your workflowr project makes it easier to run your code on other machines.

Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility.

The results in this page were generated with repository version 0d91238. See the Past versions tab to see a history of the changes made to the R Markdown and HTML files.

Note that you need to be careful to ensure that all relevant files for the analysis have been committed to Git prior to generating the results (you can use wflow_publish or wflow_git_commit). workflowr only checks the R Markdown file, but you know if there are other scripts or data files that it depends on. Below is the status of the Git repository when the results were generated:


Ignored files:
    Ignored:    .Rhistory
    Ignored:    .Rproj.user/
    Ignored:    .Ruserdata/
    Ignored:    .drake/
    Ignored:    .vscode/
    Ignored:    analysis/rosm.cache/
    Ignored:    data/
    Ignored:    inst/Benchmark for Filter Methods for Feature Selection in High-Dimensional  Classification Data.pdf
    Ignored:    inst/study-area-map/._study-area.qgs
    Ignored:    inst/study-area-map/study-area.qgs~
    Ignored:    log/
    Ignored:    renv/library/
    Ignored:    renv/local/
    Ignored:    renv/staging/
    Ignored:    reviews/
    Ignored:    rosm.cache/

Unstaged changes:
    Modified:   _drake.R
    Modified:   code/070-benchmark-matrix.R

Note that any generated files, e.g. HTML, png, CSS, etc., are not included in this status report because it is ok for generated content to have uncommitted changes.


These are the previous versions of the repository in which changes were made to the R Markdown (analysis/filter-correlation.Rmd) and HTML (docs/filter-correlation.html) files. If you’ve configured a remote Git repository (see ?wflow_git_remote), click on the hyperlinks in the table below to view the files as they were in that past version.

File Version Author Date Message
html 0d91238 pat-s 2021-11-02 Build site.
html 6c153bc pat-s 2021-04-05 Build site.
Rmd 295d836 pat-s 2021-04-05 wflow_publish(knitr_in(“analysis/filter-correlation.Rmd”), view = FALSE,
html 5ed8cbe pat-s 2021-02-03 Build site.
html a953299 pat-s 2020-08-12 Build site.
html 8b5e422 pat-s 2020-08-05 Build site.
html 3b79fd7 pat-s 2020-03-03 Build site.
Rmd f26b9f1 pat-s 2020-03-03 wflow_publish(knitr_in(“analysis/filter-correlation.Rmd”), view =
html 93feaa4 pat-s 2020-02-25 Build site.
html 1054020 pat-s 2020-02-17 Build site.
Rmd 8f344fa pat-s 2020-02-17 wflow_publish(knitr_in(“analysis/filter-correlation.Rmd”), view =
html 25f06fc pat-s 2020-01-15 Build site.
Rmd 6210a04 pat-s 2020-01-15 wflow_publish(knitr_in(“analysis/filter-correlation.Rmd”), view =
html 7fae583 pat-s 2019-12-21 Build site.
html 27d4ac5 pat-s 2019-09-02 Build site.
Rmd 518d0cb pat-s 2019-09-01 style files using tidyverse style
html 7582c67 pat-s 2019-08-31 Build site.
html f9e682f pat-s 2019-08-29 Build site.
html dc1a55d pat-s 2019-08-12 Build site.
html 9a41e71 pat-s 2019-08-06 add defoliation images
html df85aba pat-s 2019-07-12 Build site.
html 3a44a95 pat-s 2019-07-10 Build site.
html 869c409 pat-s 2019-07-02 Build site.
Rmd 24e318f pat-s 2019-07-01 update reports
html 09f6292 pat-s 2019-06-30 Build site.
Rmd 824677a pat-s 2019-06-30 workflowr::wflow_publish(“analysis/filter-correlation.Rmd”)
Rmd aff143c pat-s 2019-06-28 export filter-correlation images
html e81f421 pat-s 2019-06-27 Build site.
Rmd 644b1d8 pat-s 2019-06-27 wflow_publish(knitr_in(“analysis/filter-correlation.Rmd”), view =
html 36d883b pat-s 2019-06-27 Build site.
Rmd 92a8a9d pat-s 2019-06-27 wflow_publish(knitr_in(“analysis/filter-correlation.Rmd”), view =
html db3955e pat-s 2019-06-27 Build site.
Rmd ca7205f pat-s 2019-06-27 add new report
html 2a058f1 pat-s 2019-06-26 Build site.
Rmd acfd762 pat-s 2019-06-26 wflow_publish(knitr_in(“analysis/filter-correlation.Rmd”), view =

Correlation of filter methods

Spearman’s rank correlation is used because rankings are compared.

VI

Filter methods amongst each other

The idea behind is was to analyze the correlation between filter rankings. We only wanted to included filters which have a somewhat unique ranking. Otherwise, when creating ensemble filters, certain filters would implicitly be weighted more than others.

Takeaway:

  • Only use one of “information gain”, “gain ratio”, “sym uncert”

  • Either use Spearman or Pearson correlation

Version Author Date
0d91238 pat-s 2021-11-02
6c153bc pat-s 2021-04-05
8b5e422 pat-s 2020-08-05
1054020 pat-s 2020-02-17
25f06fc pat-s 2020-01-15
7fae583 pat-s 2019-12-21
27d4ac5 pat-s 2019-09-02
7582c67 pat-s 2019-08-31
f9e682f pat-s 2019-08-29
dc1a55d pat-s 2019-08-12
3a44a95 pat-s 2019-07-10
869c409 pat-s 2019-07-02
09f6292 pat-s 2019-06-30

NRI

Version Author Date
0d91238 pat-s 2021-11-02
6c153bc pat-s 2021-04-05
8b5e422 pat-s 2020-08-05
1054020 pat-s 2020-02-17
25f06fc pat-s 2020-01-15
7fae583 pat-s 2019-12-21
27d4ac5 pat-s 2019-09-02
7582c67 pat-s 2019-08-31
f9e682f pat-s 2019-08-29
dc1a55d pat-s 2019-08-12
3a44a95 pat-s 2019-07-10
869c409 pat-s 2019-07-02

HR

Version Author Date
0d91238 pat-s 2021-11-02
6c153bc pat-s 2021-04-05
8b5e422 pat-s 2020-08-05
1054020 pat-s 2020-02-17
25f06fc pat-s 2020-01-15
7fae583 pat-s 2019-12-21
27d4ac5 pat-s 2019-09-02
7582c67 pat-s 2019-08-31
f9e682f pat-s 2019-08-29
dc1a55d pat-s 2019-08-12
3a44a95 pat-s 2019-07-10
869c409 pat-s 2019-07-02
09f6292 pat-s 2019-06-30

Number of bins of FSelectorRcpp::information.gain()

Analyzing the effect of a different nbins value on the filter values of filter “Information Gain”.

  • Lower correlation / highest difference: nbins = 5 vs. nbins = 30

-> We decided to use with nbins = 10 in the analysis.

The hidden default of nbins when setting equal = TRUE in FSelectorRcpp::information_gain() is 5.

Version Author Date
6c153bc pat-s 2021-04-05
8b5e422 pat-s 2020-08-05
1054020 pat-s 2020-02-17
7fae583 pat-s 2019-12-21
09f6292 pat-s 2019-06-30

R version 4.0.4 (2021-02-15)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS Linux 7 (Core)

Matrix products: default
BLAS:   /opt/spack/opt/spack/linux-centos7-x86_64/gcc-9.2.0/r-4.0.4-udi7a3ahhtokdcoyqdbndhebeupt7hid/rlib/R/lib/libRblas.so
LAPACK: /opt/spack/opt/spack/linux-centos7-x86_64/gcc-9.2.0/r-4.0.4-udi7a3ahhtokdcoyqdbndhebeupt7hid/rlib/R/lib/libRlapack.so

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices datasets  utils     methods   base     

other attached packages:
 [1] fs_1.5.0         glue_1.4.2       purrr_0.3.4      ggcorrplot_0.1.3
 [5] ggplot2_3.3.3    tidyr_1.1.2      dplyr_1.0.4      usethis_2.0.0   
 [9] magrittr_2.0.1   drake_7.13.2    

loaded via a namespace (and not attached):
 [1] storr_1.2.5       progress_1.2.2    tidyselect_1.1.0  xfun_0.20        
 [5] reshape2_1.4.4    colorspace_2.0-0  vctrs_0.3.6       generics_0.1.0   
 [9] htmltools_0.5.1.1 yaml_2.2.1        rlang_0.4.10      R.oo_1.24.0      
[13] later_1.1.0.1     pillar_1.4.7      txtq_0.2.3        withr_2.4.1      
[17] R.utils_2.10.1    plyr_1.8.6        lifecycle_0.2.0   stringr_1.4.0    
[21] munsell_0.5.0     gtable_0.3.0      workflowr_1.6.2   R.methodsS3_1.8.1
[25] evaluate_0.14     labeling_0.4.2    knitr_1.31        httpuv_1.5.5     
[29] parallel_4.0.4    highr_0.8         Rcpp_1.0.6        renv_0.13.2      
[33] backports_1.2.1   promises_1.1.1    scales_1.1.1      filelock_1.0.2   
[37] farver_2.0.3      hms_1.0.0         digest_0.6.27     stringi_1.5.3    
[41] rprojroot_2.0.2   grid_4.0.4        cli_2.4.0         tools_4.0.4      
[45] base64url_1.4     tibble_3.0.6      crayon_1.4.0      whisker_0.4      
[49] pkgconfig_2.0.3   ellipsis_0.3.1    data.table_1.13.6 prettyunits_1.1.1
[53] rmarkdown_2.6     rstudioapi_0.13   R6_2.5.0          igraph_1.2.6     
[57] compiler_4.0.4    git2r_0.28.0