EVoC clustering orchestrator (ports clustering.evoc_clusters + EVoC.fit_predict). Pipeline: cosine kNN -> fuzzy simplicial set (umap-rstr) -> node embedding -> multi-resolution cluster layers; labels_ = argmax(persistence).
The numeric kernels are deftm (embedding, barcode, persistence, MST/tree, graph) so they devirtualize / fit raster's compile path; only the wiring is plain defn.
EVoC clustering orchestrator (ports clustering.evoc_clusters + EVoC.fit_predict). Pipeline: cosine kNN -> fuzzy simplicial set (umap-rstr) -> node embedding -> multi-resolution cluster layers; labels_ = argmax(persistence). The numeric kernels are deftm (embedding, barcode, persistence, MST/tree, graph) so they devirtualize / fit raster's compile path; only the wiring is plain defn.
(->coo->csr!_m_ints_ints_doubles_long_long_ints_ints_doubles_ints_Impl)Positional factory function for class evoc.coo->csr!_m_ints_ints_doubles_long_long_ints_ints_doubles_ints_Impl.
Positional factory function for class evoc.coo->csr!_m_ints_ints_doubles_long_long_ints_ints_doubles_ints_Impl.
(->make-epochs-per-sample!_m_doubles_long_long_doubles_Impl)Positional factory function for class evoc.make-epochs-per-sample!_m_doubles_long_long_doubles_Impl.
Positional factory function for class evoc.make-epochs-per-sample!_m_doubles_long_long_doubles_Impl.
(->scale-into_m_doubles_long_double_doubles_Impl)Positional factory function for class evoc.scale-into_m_doubles_long_double_doubles_Impl.
Positional factory function for class evoc.scale-into_m_doubles_long_double_doubles_Impl.
(coo->csr!_m_ints_ints_doubles_long_long_ints_ints_doubles_ints & args__68196)(fit-predict X
n
dim
&
{:keys [k n-epochs min-samples base-min-cluster-size noise seed
max-layers min-similarity]
:or {k 15
n-epochs 50
min-samples 5
base-min-cluster-size 5
noise 0.5
seed 42
max-layers 10
min-similarity 0.2}})Cluster X (flat row-major double[n*dim]). Returns {:labels int[n] (noise=-1) :persistence [..] :layers [int[]] :embedding double[n*out-dim]}. Mirrors evoc.EVoC().fit_predict with a deterministic PCA init.
Cluster X (flat row-major double[n*dim]). Returns {:labels int[n] (noise=-1)
:persistence [..] :layers [int[]] :embedding double[n*out-dim]}. Mirrors
evoc.EVoC().fit_predict with a deterministic PCA init.(make-epochs-per-sample!_m_doubles_long_long_doubles & args__68144)(scale-into_m_doubles_long_double_doubles & args__68248)cljdoc builds & hosts documentation for Clojure/Script libraries
| Ctrl+k | Jump to recent docs |
| ← | Move to previous article |
| → | Move to next article |
| Ctrl+/ | Jump to the search field |