You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Jan 8, 2026. It is now read-only.
Copy file name to clipboardExpand all lines: docs/source/methodological_framework.rst
+4-4Lines changed: 4 additions & 4 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -2,12 +2,12 @@
2
2
Methodological Framework
3
3
=========================
4
4
5
-
This document describes the methodological decisions and theoretical considerations underlying the ``clusx`` implementation. It explains how key algorithms are implemented, why specific approaches were chosen, and their academic foundations.
5
+
This document describes the methodological decisions and theoretical considerations underlying the Clusterium implementation. It explains how key algorithms are implemented, why specific approaches were chosen, and their academic foundations.
6
6
7
7
Evaluation methodology
8
8
----------------------
9
9
10
-
This section documents the methodological considerations behind the evaluation metrics implemented in ``clusx``.
10
+
This section documents the methodological considerations behind the evaluation metrics implemented in Clusterium.
11
11
12
12
Silhouette Score Calculation
13
13
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
@@ -24,7 +24,7 @@ where :math:`a(i)` is the mean distance between sample *i* and all other samples
24
24
25
25
**Our approach:**
26
26
27
-
Rather than returning a zero score when any singleton clusters exist (which would effectively discard valuable information about well-formed clusters), ``clusx`` implements a more nuanced
27
+
Rather than returning a zero score when any singleton clusters exist (which would effectively discard valuable information about well-formed clusters), Clusterium implements a more nuanced
28
28
approach that:
29
29
30
30
1. Identifies valid clusters (those with ≥2 samples)
@@ -54,7 +54,7 @@ Consider a clustering result with 10 samples and 3 clusters:
54
54
55
55
Without filtering, the silhouette score would be zero because of the singleton cluster.
56
56
57
-
With filtering, ``clusx`` would:
57
+
With filtering, Clusterium would:
58
58
59
59
1. Identify the 2 valid clusters (Cluster 1 and Cluster 3)
60
60
2. Filter the samples to include only those in valid clusters
0 commit comments