From 11b57d4af8d81dfa8f98ba795787feb613604b43 Mon Sep 17 00:00:00 2001
From: clark-hive <clark@thehive.ai>
Date: Fri, 9 Sep 2022 02:36:20 +0000
Subject: [PATCH] working around issue in kmc2

---
 python/bolt/bolt_api.py | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/python/bolt/bolt_api.py b/python/bolt/bolt_api.py
index 291119c5..5bf1d898 100644
--- a/python/bolt/bolt_api.py
+++ b/python/bolt/bolt_api.py
@@ -137,6 +137,9 @@ def _learn_centroids(X, ncentroids, ncodebooks):
         start_col = i * subvect_len
         end_col = start_col + subvect_len
         X_in = X[:, start_col:end_col]
+        if np.unique(X_in).size == 1 and X_in.size > 1:
+					#kmc2 has a bug if all rows of matrix are duplicates
+          X_in[0,0] += 0.001
         centroids, labels = kmeans(X_in, ncentroids)
         ret[:, i, :] = centroids