more

.devel/pytest/test_approx.py

@@ -19,11 +19,11 @@
     #rpy2.robjects.numpy2ri.activate()
     stats = importr("stats")
 
-    r_base = importr("base")
-    lib_loc = r_base.Sys_getenv("R_LIBS_USER")[0]
-    print(lib_loc)
+    # r_base = importr("base")
+    # lib_loc = r_base.Sys_getenv("R_LIBS_USER")[0]
+    # print(lib_loc)
 
-    genie = importr("genie", lib_loc=lib_loc)
+    genie = importr("genie")  #, lib_loc=lib_loc)
 except:
     rpy2 = None
     stats = None

.devel/sphinx/index.md

@@ -57,7 +57,7 @@ pip3 install genieclust
 ```
 
 from the command line or through your favourite package manager.
-Note a familiar *scikit-learn*-like {cite}`sklearn_api` look-and-feel:
+Note the *scikit-learn*-like {cite}`sklearn_api` API:
 
 ```python
 import genieclust
@@ -73,6 +73,7 @@ labels = g.fit_predict(X)
 ::::
 
 
+
 ## R Version
 
 The **R version** of *genieclust* can be downloaded from

.devel/sphinx/weave/basics.Rmd

@@ -1,6 +1,5 @@
 # Basics
 
-
 *Genie* {cite}`genieins` is an agglomerative hierarchical clustering
 algorithm that links clusters minding that
 the Gini index (a measure of inequality used in, amongst others,
@@ -19,7 +18,6 @@ via a call to `pip3 install genieclust` from the command line.
 
 
 
-
 ```{python imports,results="hide"}
 import numpy as np
 import pandas as pd
@@ -72,9 +70,8 @@ plt.show()
 ```
 
 Let us apply the Genie algorithm (with the default/recommended
-`gini_threshold` parameter value). The `genieclust` package's interface
-is compatible with the one from the popular
-[scikit-learn](https://scikit-learn.org/) library {cite}`sklearn`.
+`gini_threshold` parameter value). The `genieclust` package's programming
+interface is [scikit-learn](https://scikit-learn.org/)-compatible {cite}`sklearn`.
 In particular, an object of class `Genie` is equipped with the
 `fit` and `fit_predict` methods {cite}`sklearn_api`.
 
@@ -159,8 +156,6 @@ genieclust.compare_partitions.adjusted_rand_score(labels_true, labels_kmeans)
 The adjusted Rand score of $\sim 0.3$ indicates a far-from-perfect fit.
 
 
-
-
 ## A Comparison with HDBSCAN\*
 
 Let's also make a comparison against a version of the DBSCAN
@@ -170,7 +165,6 @@ the [hdbscan](https://github.com/scikit-learn-contrib/hdbscan) package
 {cite}`hdbscanpkg` implements its robustified variant
 {cite}`hdbscan`, which makes the algorithm much more user-friendly.
 
-
 Here are the clustering results with the `min_cluster_size` parameter
 of 3, 5, 10, and 15:
 
@@ -247,7 +241,7 @@ plt.show()
 
 Dendrogram plotting is possible with `scipy.cluster.hierarchy`:
 
-```{python basics-dendrogram-1,fig.cap="Example dendrogram."}
+```{python basics-dendrogram-1,fig.cap="Example dendrogram.",results="hide"}
 import scipy.cluster.hierarchy
 g = genieclust.Genie(compute_full_tree=True)
 g.fit(X)
@@ -259,7 +253,7 @@ plt.show()
 
 Another example:
 
-```{python basics-dendrogram-2,fig.cap="Another example dendrogram."}
+```{python basics-dendrogram-2,fig.cap="Another example dendrogram.",results="hide"}
 scipy.cluster.hierarchy.dendrogram(linkage_matrix,
     truncate_mode="lastp", p=15, orientation="left")
 plt.show()

.devel/sphinx/weave/benchmarks_ar.Rmd

@@ -30,8 +30,6 @@ For more detailed results based on other partition similarity scores,
 see the [Appendix](benchmarks_details).
 
 
-
-
 ```{python imports,results="hide",echo=FALSE}
 import numpy as np
 import pandas as pd
@@ -185,7 +183,6 @@ res_max = res2.groupby(["dataset", "method"]).max().\
 
 
 
-
 ```{python plot_small,echo=FALSE,results="hide",warn=FALSE,fig.cap="Distribution of the AR index for each algorithm (small datasets); best=1.0.",fig.height=5.9375}
 do_plot(res_max)
 ```
@@ -199,8 +196,6 @@ compare {cite}`cvimst`)
 tend to output good quality outcomes as well.
 
 
-
-
 Descriptive statistics for the ranks (for each dataset,
 each algorithm that gets the highest AR index rounded to 2 decimal digits,
 gets a rank of 1); lower ranks are better:

.devel/sphinx/weave/benchmarks_ar.md

@@ -37,8 +37,6 @@ see the [Appendix](benchmarks_details).
 
 
 
-
-
 ## Small Datasets
 
 As some of the algorithms tested here have failed to generate a solution
@@ -62,7 +60,6 @@ Moreover, Gaussian mixtures used `n_init=100`.
 
 
 
-
 (fig:plot_small)=
 ```{figure} benchmarks_ar-figures/plot_small-1.*
 Distribution of the AR index for each algorithm (small datasets); best=1.0.
@@ -77,8 +74,6 @@ compare {cite}`cvimst`)
 tend to output good quality outcomes as well.
 
 
-
-
 Descriptive statistics for the ranks (for each dataset,
 each algorithm that gets the highest AR index rounded to 2 decimal digits,
 gets a rank of 1); lower ranks are better:
@@ -96,7 +91,7 @@ gets a rank of 1); lower ranks are better:
 | K-means           |      72 |    5.6 |   3.8 |     1 |   1   |     6 |   9   |    12 |
 | Single linkage    |      72 |    7.4 |   5.1 |     1 |   1   |    11 |  12   |    12 |
 | Spectral_RBF_5    |      72 |    5.2 |   3.5 |     1 |   1   |     6 |   8   |    11 |
-| Ward linkage      |      72 |    6   |   3   |     1 |   4   |     6 |   8   |    12 |
+| Ward linkage      |      72 |    6   |   3   |     1 |   4   |     6 |   8   |    12 | 
 
 
 ## Large Datasets
@@ -127,7 +122,7 @@ Descriptive statistics for the ranks (AR index):
 | ITM            |       6 |    3.3 |   2.3 |     1 |   1.5 |   3   |   5.2 |     6 |
 | K-means        |       6 |    3.3 |   1.6 |     1 |   2.2 |   3.5 |   4.8 |     5 |
 | Single linkage |       6 |    6.8 |   0.4 |     6 |   7   |   7   |   7   |     7 |
-| Ward linkage   |       6 |    3.2 |   1.5 |     1 |   2.2 |   3.5 |   4   |     5 |
+| Ward linkage   |       6 |    3.2 |   1.5 |     1 |   2.2 |   3.5 |   4   |     5 | 
 
 
 

.devel/sphinx/weave/noise.md

@@ -5,8 +5,7 @@
 # Clustering with Noise Points Detection
 
 
-
-```python
+``` python
 import numpy as np
 import pandas as pd
 import matplotlib.pyplot as plt
@@ -22,8 +21,7 @@ the at [hdbscan](https://github.com/scikit-learn-contrib/hdbscan)
 {cite}`hdbscanpkg` package's project site:
 
 
-
-```python
+``` python
 dataset = "hdbscan"
 X = np.loadtxt("%s.data.gz" % dataset, ndmin=2)
 labels_true = np.loadtxt("%s.labels0.gz" % dataset, dtype=np.intp) - 1
@@ -38,11 +36,13 @@ find useful for whatever their goal is).
 The `-1` labels denote noise points (light grey markers).
 
 
-
-```python
+``` python
 genieclust.plots.plot_scatter(X, labels=labels_true, alpha=0.5)
 plt.title("(n=%d, true n_clusters=%d)" % (X.shape[0], n_clusters))
 plt.axis("equal")
+```
+
+``` python
 plt.show()
 ```
 
@@ -66,8 +66,7 @@ Here are the effects of playing with the `M` parameter
 (we keep the default `gini_threshold`):
 
 
-
-```python
+``` python
 Ms = [2, 5, 10, 25]
 for i in range(len(Ms)):
     g = genieclust.Genie(n_clusters=n_clusters, M=Ms[i])
@@ -76,6 +75,9 @@ for i in range(len(Ms)):
     genieclust.plots.plot_scatter(X, labels=labels_genie, alpha=0.5)
     plt.title("(gini_threshold=%g, M=%d)"%(g.gini_threshold, g.M))
     plt.axis("equal")
+```
+
+``` python
 plt.show()
 ```
 
@@ -91,8 +93,7 @@ and then apply the clustering procedure once again
 but now with respect to the original distance (here: Euclidean):
 
 
-
-```python
+``` python
 # Step 1: Noise point identification
 g1 = genieclust.Genie(n_clusters=n_clusters, M=50)
 labels_noise = g1.fit_predict(X)
@@ -106,6 +107,9 @@ labels_noise[non_noise] = labels_genie
 genieclust.plots.plot_scatter(X, labels=labels_noise, alpha=0.5)
 plt.title("(gini_threshold=%g, noise points removed first; M=%d)"%(g2.gini_threshold, g1.M))
 plt.axis("equal")
+```
+
+``` python
 plt.show()
 ```
 
@@ -126,8 +130,7 @@ of finer or coarser granularity.
 
 
 
-
-```python
+``` python
 ncs = [5, 6, 7, 8, 10, 15]
 for i in range(len(ncs)):
     g = genieclust.Genie(n_clusters=ncs[i])
@@ -137,6 +140,9 @@ for i in range(len(ncs)):
     genieclust.plots.plot_scatter(X, labels=labels_noise, alpha=0.5)
     plt.title("(n_clusters=%d)"%(g.n_clusters))
     plt.axis("equal")
+```
+
+``` python
 plt.show()
 ```
 
@@ -153,15 +159,21 @@ Labels predicted by Genie when noise points were removed from the dataset – di
 Here are the results returned by `hdbscan` with default parameters:
 
 
-
-```python
+``` python
 import hdbscan
+```
+
+
+``` python
 h = hdbscan.HDBSCAN()
 labels_hdbscan = h.fit_predict(X)
 genieclust.plots.plot_scatter(X, labels=labels_hdbscan, alpha=0.5)
 plt.title("(min_cluster_size=%d, min_samples=%d)" % (
     h.min_cluster_size, h.min_samples or h.min_cluster_size))
 plt.axis("equal")
+```
+
+``` python
 plt.show()
 ```
 
@@ -177,8 +189,7 @@ we can obtain a partition that is even closer to the reference one:
 
 
 
-
-```python
+``` python
 mcss = [5, 10, 25]
 mss = [5, 10]
 for i in range(len(mcss)):
@@ -190,6 +201,9 @@ for i in range(len(mcss)):
         plt.title("(min_cluster_size=%d, min_samples=%d)" % (
             h.min_cluster_size, h.min_samples or h.min_cluster_size))
         plt.axis("equal")
+```
+
+``` python
 plt.show()
 ```
 

.devel/sphinx/weave/r.Rmd

@@ -10,8 +10,6 @@ install.packages("genieclust")
 
 
 Below are a few basic examples of how to interact with the package.
-<!-- (partially based on Marek's [forthcoming book](https://lmlcr.gagolewski.com)). -->
-
 
 ```{r load}
 library("genieclust")

.devel/sphinx/weave/sklearn_toy_example.md

@@ -16,8 +16,7 @@ on a larger data sample and with the Genie algorithm in the game.
 
 
 
-
-```python
+``` python
 import time
 import warnings
 
@@ -41,8 +40,7 @@ First, we generate the datasets. Note that in the
 `n_samples` was set to 1500.
 
 
-
-```python
+``` python
 n_samples = 10000
 noisy_circles = datasets.make_circles(n_samples=n_samples, factor=.5,
                                       noise=.05)
@@ -66,8 +64,7 @@ varied = datasets.make_blobs(n_samples=n_samples,
 Then, we run the clustering procedures and plot the results.
 
 
-
-```python
+``` python
 # Set up cluster parameters
 plt.figure(figsize=(9 * 1.3 + 2, 14.5))
 plt.subplots_adjust(left=.02, right=.98, bottom=.001, top=.96, wspace=.05,