more

Author: gagolews

.devel/sphinx/bibliography.bib

@@ -1,21 +1,23 @@
-@article{nca,
-    author = {M. Gagolewski},
-    title = {Normalised clustering accuracy: {A}n asymmetric external cluster validity measure},
-    journal = {Journal of Classification},
-    year = {2024},
-    url = {https://link.springer.com/content/pdf/10.1007/s00357-024-09482-2.pdf},
-    doi = {10.1007/s00357-024-09482-2},
-    note = {in press}
-}
-
 @article{cvimst,
     author = {M. Gagolewski and A. Cena and M. Bartoszuk and L. Brzozowski},
     title = {Clustering with minimum spanning trees: {H}ow good can it be?},
     journal = {Journal of Classification},
-    year = {2024},
+    year = {2025},
+    volume = {42},
+    pages = {90--112},
     url = {https://link.springer.com/content/pdf/10.1007/s00357-024-09483-1.pdf},
     doi = {10.1007/s00357-024-09483-1},
-    note = {in press}
+}
+
+@article{nca,
+    author = {M. Gagolewski},
+    title = {Normalised clustering accuracy: {A}n asymmetric external cluster validity measure},
+    journal = {Journal of Classification},
+    year = {2025},
+    volume = {42},
+    pages = {2--30},
+    url = {https://link.springer.com/content/pdf/10.1007/s00357-024-09482-2.pdf},
+    doi = {10.1007/s00357-024-09482-2},
 }
 
 @article{clustering_benchmarks,

.devel/sphinx/weave/data-v1.Rmd

@@ -90,3 +90,10 @@ cat(readLines("include-dataset-browser.js"), sep="\n")
 ```
 </script>
 ::::
+
+
+::::{important}
+As a courtesy, **please cite** the original source as well as the current project
+{cite}`clustering_benchmarks` as well as mention {cite}`clustering_data_v1`
+which gives the exact version and URL of the dataset suite. Thank you.
+::::

.devel/sphinx/weave/data-v1.md

@@ -217,3 +217,10 @@ window.onhashchange = locationHashChanged;
 locationHashChanged();
 </script>
 ::::
+
+
+::::{important}
+As a courtesy, **please cite** the original source as well as the current project
+{cite}`clustering_benchmarks` as well as mention {cite}`clustering_data_v1`
+which gives the exact version and URL of the dataset suite. Thank you.
+::::

.devel/sphinx/weave/how-to-access.Rmd

@@ -113,9 +113,10 @@ former can be called from within the latter.
 
 ## Julia
 
-Very similar to Python and R the datasets can be accessed in Julia using the [*CSV.jl*](https://csv.juliadata.org) package.
+Very similar to Python and R the datasets can be accessed
+in Julia using the [*CSV.jl*](https://csv.juliadata.org) package.
 
-```{julia}
+```julia
 using CSV
 
 base_name = joinpath("~", "Projects", "clustering-data-v1", "wut", "smile")
@@ -128,5 +129,8 @@ labels = CSV.read(base_name * ".labels0.gz", CSV.Tables.matrix; header=false, de
 ::::{todo}
 Contributions are welcome: Describe how to load
 the datasets and benchmark results
-in GNU Octave, Scilab, Julia, Mathematica, ... (🚧 help needed 🚧)
+in GNU Octave, Scilab, Mathematica, ... (🚧 help needed 🚧)
+
+Thanks to [Torsten Stöter](https://github.com/tstoeter) for contributing
+the Julia code.
 ::::

.devel/sphinx/weave/how-to-access.md

@@ -135,9 +135,27 @@ former can be called from within the latter.
 
 
 
+## Julia
+
+Very similar to Python and R the datasets can be accessed
+in Julia using the [*CSV.jl*](https://csv.juliadata.org) package.
+
+
+```julia
+using CSV
+
+base_name = joinpath("~", "Projects", "clustering-data-v1", "wut", "smile")
+base_name = expanduser(base_name)
+data = CSV.read(base_name * ".data.gz", CSV.Tables.matrix; header=false, delim=' ')
+labels = CSV.read(base_name * ".labels0.gz", CSV.Tables.matrix; header=false, delim=' ')
+```
+
 
 ::::{todo}
 Contributions are welcome: Describe how to load
 the datasets and benchmark results
-in GNU Octave, Scilab, Julia, Mathematica, ... (🚧 help needed 🚧)
+in GNU Octave, Scilab, Mathematica, ... (🚧 help needed 🚧)
+
+Thanks to [Torsten Stöter](https://github.com/tstoeter) for contributing
+the Julia code.
 ::::

.devel/sphinx/weave/suite-v1.Rmd

@@ -44,7 +44,7 @@ def summarise_battery(battery):
 (sec:suite-v1)=
 # Benchmark Suite (v`r VERSION`)
 
-We have compiled **a large suite of benchmark datasets**.
+We have compiled, curated, and polished **a large suite of benchmark datasets**.
 For reproducibility, the datasets and label vectors are **versioned**.
 
 
@@ -77,7 +77,7 @@ index *g*, where $g=0$ means that all clusters consist of the same number
 of points.
 
 
-::::{important}
+::::{note}
 The versioned **snapshots of the suite** are available for download at:
 <https://github.com/gagolews/clustering-data-v1/releases/tag/v`r VERSION`>.
 
@@ -107,9 +107,11 @@ each dataset is accompanied by a text file specifying more details thereon
 (e.g., the literature references that we are asked to cite).
 
 
-As a courtesy, **please cite** also the current project
+::::{important}
+As a courtesy, **please cite** the original source as well as the current project
 {cite}`clustering_benchmarks` as well as mention {cite}`clustering_data_v1`
 which gives the exact version and URL of the dataset suite. Thank you.
+::::
 
 
 There is some inherent overlap between the original databases.
@@ -189,7 +191,7 @@ summarise_battery("sipu")
 (sec:battery-fcps)=
 ## `fcps`
 
-9 datasets from the *Fundamental Clustering Problem Suite*
+Nine datasets from the *Fundamental Clustering Problem Suite*
 proposed by A. Ultsch {cite}`fcps` from the Marburg University,
 Germany.
 
@@ -214,7 +216,7 @@ summarise_battery("fcps")
 (sec:battery-graves)=
 ## `graves`
 
-10 *synthetic data sets* discussed by D. Graves and W. Pedrycz
+Ten *synthetic data sets* discussed by D. Graves and W. Pedrycz
 in {cite}`graves`.
 
 The dataset consist of 200–1050 observations in 2 dimensions.
@@ -272,7 +274,7 @@ summarise_battery("other")
 (sec:battery-uci)=
 ## `uci`
 
-A selection of 8 high-dimensional datasets available through the UCI
+A selection of eight high-dimensional datasets available through the UCI
 (University of California, Irvine)
 [Machine Learning Repository](http://archive.ics.uci.edu/ml/) {cite}`uci`.
 Some of them were considered for benchmark purposes

.devel/sphinx/weave/suite-v1.md

@@ -10,7 +10,7 @@
 (sec:suite-v1)=
 # Benchmark Suite (v1.1.0)
 
-We have compiled **a large suite of benchmark datasets**.
+We have compiled, curated, and polished **a large suite of benchmark datasets**.
 For reproducibility, the datasets and label vectors are **versioned**.
 
 
@@ -43,7 +43,7 @@ index *g*, where $g=0$ means that all clusters consist of the same number
 of points.
 
 
-::::{important}
+::::{note}
 The versioned **snapshots of the suite** are available for download at:
 <https://github.com/gagolews/clustering-data-v1/releases/tag/v1.1.0>.
 
@@ -70,9 +70,11 @@ each dataset is accompanied by a text file specifying more details thereon
 (e.g., the literature references that we are asked to cite).
 
 
-As a courtesy, **please cite** also the current project
+::::{important}
+As a courtesy, **please cite** the original source as well as the current project
 {cite}`clustering_benchmarks` as well as mention {cite}`clustering_data_v1`
 which gives the exact version and URL of the dataset suite. Thank you.
+::::
 
 
 There is some inherent overlap between the original databases.
@@ -203,7 +205,7 @@ We excluded the `DIM`-sets as they are too easy for most algorithms.
 (sec:battery-fcps)=
 ## `fcps`
 
-9 datasets from the *Fundamental Clustering Problem Suite*
+Nine datasets from the *Fundamental Clustering Problem Suite*
 proposed by A. Ultsch {cite}`fcps` from the Marburg University,
 Germany.
 
@@ -238,7 +240,7 @@ see also {cite}`ThrunUltsch2020:fcps`.
 (sec:battery-graves)=
 ## `graves`
 
-10 *synthetic data sets* discussed by D. Graves and W. Pedrycz
+Ten *synthetic data sets* discussed by D. Graves and W. Pedrycz
 in {cite}`graves`.
 
 The dataset consist of 200–1050 observations in 2 dimensions.
@@ -316,7 +318,7 @@ Datasets from multiple sources:
 (sec:battery-uci)=
 ## `uci`
 
-A selection of 8 high-dimensional datasets available through the UCI
+A selection of eight high-dimensional datasets available through the UCI
 (University of California, Irvine)
 [Machine Learning Repository](http://archive.ics.uci.edu/ml/) {cite}`uci`.
 Some of them were considered for benchmark purposes

docs/clustbench-documentation.html

@@ -895,7 +895,7 @@ <h1>Documentation<a class="headerlink" href="#documentation" title="Link to this
               Some rights reserved. Licensed under <a href='https://creativecommons.org/licenses/by-nc-nd/4.0/'>CC BY-NC-ND 4.0</a>.
               Built with <a href="https://sphinx-doc.org/">Sphinx</a>
               and a customised <a href="https://github.com/pradyunsg/furo">Furo</a> theme.
-              Last updated on 2025-04-08T13:43:38+0200.
+              Last updated on 2025-05-21T11:51:24+0200.
               This site will never display any ads: it is a non-profit project.
               It does not collect any data.
             </div>

docs/genindex.html

@@ -452,7 +452,7 @@ <h2>T</h2>
               Some rights reserved. Licensed under <a href='https://creativecommons.org/licenses/by-nc-nd/4.0/'>CC BY-NC-ND 4.0</a>.
               Built with <a href="https://sphinx-doc.org/">Sphinx</a>
               and a customised <a href="https://github.com/pradyunsg/furo">Furo</a> theme.
-              Last updated on 2025-04-08T13:43:38+0200.
+              Last updated on 2025-05-21T11:51:24+0200.
               This site will never display any ads: it is a non-profit project.
               It does not collect any data.
             </div>

docs/index.html

@@ -431,7 +431,7 @@ <h1>A Framework for Benchmarking Clustering Algorithms<a class="headerlink" href
               Some rights reserved. Licensed under <a href='https://creativecommons.org/licenses/by-nc-nd/4.0/'>CC BY-NC-ND 4.0</a>.
               Built with <a href="https://sphinx-doc.org/">Sphinx</a>
               and a customised <a href="https://github.com/pradyunsg/furo">Furo</a> theme.
-              Last updated on 2025-04-08T13:43:38+0200.
+              Last updated on 2025-05-21T11:51:24+0200.
               This site will never display any ads: it is a non-profit project.
               It does not collect any data.
             </div>