Apple-2-SW
/
ii-pix
mirror of https://github.com/KrisKennaway/ii-pix.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

k-means should be using median with L1 norm, otherwise it may not converge 

Also optimize a tiny bit
This commit is contained in:
kris 2021-11-13 16:10:33 +00:00
parent 5cab854269
commit 52af982159
1 changed files with 6 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
dither.pyx
View File

@ -464,7 +464,7 @@ def k_means_with_fixed_centroids(
cdef int i, iteration, centroid_idx, num_fixed_centroids, num_random_centroids, best_centroid_idx
cdef float[::1] point, centroid, new_centroid, old_centroid
cdef float[:, ::1] centroids
cdef float best_dist, centroid_movement
cdef float best_dist, centroid_movement, dist
centroids = np.zeros((n_clusters, 3), dtype=np.float32)
if fixed_centroids is not None:
@ -473,9 +473,9 @@ def k_means_with_fixed_centroids(
num_random_centroids = n_clusters - num_fixed_centroids
# TODO: kmeans++ initialization
cdef int rand_idx = random.randint(0, data.shape[0])
for i in range(num_random_centroids):
centroids[num_fixed_centroids + i, :] = data[
random.randint(0, data.shape[0]), :]
centroids[num_fixed_centroids + i, :] = data[rand_idx, :]
cdef int[::1] centroid_weights = np.zeros(n_clusters, dtype=np.int32)
for iteration in range(iterations):
@ -484,7 +484,8 @@ def k_means_with_fixed_centroids(
for point in data:
best_dist = 1e9
best_centroid_idx = 0
for centroid_idx, centroid in enumerate(centroids):
for centroid_idx in range(n_clusters):
centroid = centroids[centroid_idx, :]
dist = colour_distance(centroid, point)
if dist < best_dist:
best_dist = dist
@ -492,12 +493,11 @@ def k_means_with_fixed_centroids(
closest_points[best_centroid_idx].append(point)
centroid_movement = 0
for centroid_idx, points in closest_points.items():
centroid_weights[centroid_idx] = len(points)
if centroid_idx < num_fixed_centroids:
continue
new_centroid = np.mean(np.array(points), axis=0)
new_centroid = np.median(np.array(points), axis=0)
old_centroid = centroids[centroid_idx]
centroid_movement += colour_distance(old_centroid, new_centroid)
centroids[centroid_idx, :] = new_centroid