jketreno/ketr.photos
1
0
Fork 0
Code Issues 1 Activity

Neighbors get added to the live Set during blob expansion

Browse Source 
Signed-off-by: James Ketrenos <james_git@ketrenos.com>
This commit is contained in:
James Ketr 2023-01-17 10:14:24 -08:00
parent 50ddb96e74
commit 2800430817
1 changed files with 21 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
ketrface/ketrface/dbscan.py
View File

@ -8,13 +8,13 @@ Undefined = 0
Edge = -1 Edge = -1
Noise = -2 Noise = -2
# Union of two lists of dicts # Union of two lists of dicts, adding unique elements of B to
# end of A
def Union(A, B): def Union(A, B):
C = A
for key in B: for key in B:
if key not in C: if key not in A:
C.append(key) A.append(key)
return C return A
# https://en.wikipedia.org/wiki/DBSCAN # https://en.wikipedia.org/wiki/DBSCAN
# #
@ -53,32 +53,39 @@ def DBSCAN(points, eps = MAX_DISTANCE, minPts = MIN_PTS, verbose = True):
S.remove(P) S.remove(P)
sub_perc = -1 sub_perc = -1
sub_total = len(S) sub_last = 0
T = S.copy()
for j, Q in enumerate(S): # Process every seed point for j, Q in enumerate(S): # Process every seed point
if verbose == True: if verbose == True:
sub_total = len(S)
sub_new_perc = int(100 * (j+1) / sub_total) sub_new_perc = int(100 * (j+1) / sub_total)
now = time.time() sub_now = time.time()
if sub_new_perc != sub_perc or now - last > 5: if sub_new_perc != sub_perc or sub_now - sub_last > 5:
sub_perc = sub_new_perc sub_perc = sub_new_perc
print(f'... points {sub_perc}% ({j}/{sub_total} processed) complete with {len(clusters)} identities ({now - start}s).') print(f'... points {sub_perc}% ({j}/{sub_total} processed [{perc}% total]) complete with {len(clusters)} identities ({now - start}s).')
last = now sub_last = sub_now
if Q['cluster'] == Noise: # Change Noise to border point if Q['cluster'] == Noise: # Change Noise to border point
Q['cluster'] = C Q['cluster'] = C
C['faces'].append(Q) C['faces'].append(Q)
if Q['cluster'] != Undefined: # Previously processed (border point) if Q['cluster'] != Undefined: # Previously processed (border point)
continue continue
Q['cluster'] = C # Label neighbor Q['cluster'] = C # Label neighbor
C['faces'].append(Q) C['faces'].append(Q)
N = RangeQuery(points, Q, eps) # Find neighbors N = RangeQuery(points, Q, eps) # Find neighbors
if len(N) >= minPts: # Density check (if Q is a core point) if len(N) >= minPts: # Density check (if Q is a core point)
T = Union(T, N) # Add new neighbors to seed set S = Union(S, N) # Add new neighbors to seed set
S = T
return clusters return clusters
def RangeQuery(points, Q, eps): def RangeQuery(points, Q, eps):
neighbors = [] neighbors = []
for P in points: # Scan all points in the database for P in points: # Scan all points in the database
if P == Q:
continue
if P in neighbors: if P in neighbors:
continue continue
distance = findCosineDistance( # Compute distance and check epsilon distance = findCosineDistance( # Compute distance and check epsilon